Foundation Research — What’s Included
353-page Foundation Studies document

The following Foundation Studies and Sections provide the research and documentation underlying this report. Readers may review the sections of particular interest or consult the complete Foundation Research for detailed findings, methodology, limitations, sources, and supporting evidence.

Shakeshaft / U.S. Department of Education, 2004 — Educator Sexual Misconduct: A Synthesis of Existing Literature
Associated Press Nationwide Investigation, 2007 — Sex Abuse a Shadow Over U.S. Schools
U.S. Government Accountability Office, 2010 — K–12 Education: Selected Cases of Public and Private Schools That Hired or Retained Individuals with Histories of Sexual Misconduct
ESSA and “Passing the Trash”: Did Federal Law Fix the Problem?
Jeglic et al., 2023 — The Nature and Scope of Educator Misconduct in K–12
Grant, Haverland & Kalbfleisch — Title IX Policy Implementation and Sexual Harassment Prevalence in K–12 Schools
U.S. Department of Education — Civil Rights Data Collection: 2021–22 CRDC
Assini-Meytin et al., 2025 — Child Sexual Abuse and Boundary Violating Behaviors in Youth-Serving Organizations: National Prevalence and Distribution by Organizational Type
What Actually Prevents Sexual Abuse in Youth-Serving Organizations?
“Passing the Trash”: Why Would a School Stay Silent?
NCANDS: Can America’s Child-Protection System Tell Us How Many Children Are Maltreated by School Employees?
FBI/NIBRS: Can We Count School Employees Arrested for Sexual Offenses Against Students?
The 2007 Associated Press Investigation: What Did the Famous “2,570 Educators” Figure Actually Mean?
NASDTEC Educator Identification Clearinghouse: Could the Next School Have Known?
Building the 50-State Child-Safety Screening Matrix
Five-State Pilot: Testing the National Child-Safety Screening Matrix
The Closed-Loop Child Protection Standard: Following Information from the First Warning to the Next Hiring Decision
The 51-Jurisdiction Audit: Alabama → California
The 51-Jurisdiction Audit: Colorado • Connecticut • Delaware • Florida • Georgia
The 51-Jurisdiction Audit: Hawaii • Idaho • Illinois • Indiana • Iowa
The 51-Jurisdiction Audit: Kansas • Kentucky • Louisiana • Maine • Maryland
The 51-Jurisdiction Audit: Massachusetts • Michigan • Minnesota • Mississippi • Missouri
The 51-Jurisdiction Audit: Montana • Nebraska • Nevada • New Hampshire • New Jersey
The 51-Jurisdiction Audit: New Mexico • New York • North Carolina • North Dakota • Ohio
The 51-Jurisdiction Audit: Oklahoma • Oregon • Pennsylvania • Rhode Island • South Carolina
The 51-Jurisdiction Audit: South Dakota • Tennessee • Texas • Utah • Vermont
The 51-Jurisdiction Audit — Final Six: Virginia • Washington • West Virginia • Wisconsin • Wyoming • District of Columbia
Quality-Control / Uncertainty Audit of the 51-Jurisdiction Dataset
Quality-Control Audit — Round Two: Oklahoma • South Dakota • Montana • Massachusetts • North Carolina • Mississippi • South Carolina • Vermont • District of Columbia

The document itself skips the printed labels “Foundation Study #7” and “Foundation Section #28.” Number 7 is actually labeled “FOUNDATION SOURCE #7” in the manuscript, while the material between #27 and #29 is clearly the first quality-control/uncertainty audit but doesn’t appear to have been given a visible FOUNDATION SECTION #28 heading. So there is not discovered missing research—I’ve discovered two numbering/labeling inconsistencies we should correct in the master document.

And the progression is quite good. By #14–17, the research naturally changes from “How big is this problem?” to “Can information follow an educator through the system?” and then into our original 51-jurisdiction analysis. The manuscript itself describes NASDTEC as the interstate disciplinary-information mechanism and then asks who is actually required to use these protections.

~~~~~

Within the actual 156-page 2004 U.S. Department of Education report, there are several things in it that are even more important than the famous 9.6% statistic.

FOUNDATION STUDY #1

Shakeshaft / U.S. Department of Education, 2004

Full title: Educator Sexual Misconduct: A Synthesis of Existing Literature

Author: Charol Shakeshaft, Hofstra University
Prepared for: U.S. Department of Education, Office of the Under Secretary
Date: June 2004
Length: 156 pages

This wasn’t simply an independent paper that happened to be published by the government. Congress mandated a national study of sexual abuse in schools under §5414 of the ESEA, as amended by No Child Left Behind, and the Department contracted with Shakeshaft to review the available evidence. (ERIC)

One qualification belongs in our report from the beginning: the Department expressly said that the author’s views should not automatically be interpreted as official Department endorsement, and its preface raised concerns about the distinction between the broader term “sexual misconduct” and the legally narrower “sexual abuse.” Nevertheless, the Department called the subject critically important and said publication was in the public interest. (U.S. Department of Education)

That is exactly the kind of nuance we want.

1. What did “educator sexual misconduct” mean?

This is crucial.

Shakeshaft defined an educator much more broadly than “teacher.”

It included adults over 18 working with or for an educational organization, paid or unpaid, including teachers, counselors, administrators, secretaries, bus drivers, coaches, volunteers, lunchroom staff, tutors, music teachers and special-education aides. (U.S. Department of Education)

The conduct could be:

physical, verbal or visual.

It encompassed conduct ranging from sexual conversations/comments and displaying sexual material through sexual touching and penetration. The report ultimately defined educator sexual misconduct broadly as sexual conduct that could constitute professional misconduct, including Title IX sexual harassment, criminal sexual abuse, educator/student sexual relationships, and conduct intended to establish such a relationship. (U.S. Department of Education)

REPORT LANGUAGE

Therefore:

Educator sexual misconduct is an umbrella category. It must not be treated as synonymous with rape, sexual assault, or contact sexual abuse.

We’re going to repeat that distinction throughout the publication.

2. Where did the famous 9.6% actually come from?

This is VERY important.

Shakeshaft did not conduct a new national survey in 2004.

She performed a secondary analysis of data collected for the American Association of University Women by Harris Interactive in 2000 for Hostile Hallways.

And this was a substantial survey.

The sample was:

2,065 students in grades 8–11

drawn through a stratified two-stage design from a list of approximately 80,000 schools.

Of them:

1,559 public-school students were surveyed in school, and
505 public-school students completed surveys online.

The report states that the findings could be generalized to public-school students in grades 8–11 at the 95% confidence level, with approximately a ±4 percentage-point margin of error. (U.S. Department of Education)

The underlying AAUW study itself describes its sample as nationally representative. (ERIC)

That makes this considerably more valuable than a convenience survey.

3. What were students actually asked?

Fourteen behaviors were presented.

They ranged from sexual comments, gestures, pictures, rumors and spying through unwanted sexual touching, clothing removal, forced kissing and being forced to perform another sexual act.

And here’s an important methodological feature:

Students were initially asked about misconduct by anyone.

Then follow-up questions identified whether the offender was:

another student, teacher, counselor, other school employee, etc.

That is how Shakeshaft isolated educator misconduct from the much larger amount of student-on-student sexual misconduct. (U.S. Department of Education)

That’s a significant methodological point in the study’s favor.

4. The actual result

The secondary analysis found:

9.6%

of students in grades 8–11 reported experiencing unwanted contact and/or non-contact educator sexual misconduct during their school lives. (U.S. Department of Education)

The report further gives:

8.7% — non-contact misconduct

6.7% — contact misconduct

Those categories overlap because some students reported both. They therefore cannot be added together. (U.S. Department of Education)

And here’s another statistic worth preserving:

Among students reporting sexual misconduct in schools generally:

21% identified educators as perpetrators.

79% identified other students. (U.S. Department of Education)

That distinction is useful because it prevents us from inadvertently attributing all school sexual misconduct to employees.

5. The famous “4.5 million children” figure is an extrapolation

This one needs a big asterisk in our report.

Shakeshaft took the 9.6% finding and applied it to the total K–12 population.

That produced:

More than 4.5 million students

potentially subjected to educator sexual misconduct during their K–12 careers. (U.S. Department of Education)

But Shakeshaft explicitly said this depended upon the assumption that the AAUW results for grades 8–11 accurately represented experiences across the entire K–12 population. (U.S. Department of Education)

Therefore our report should NEVER say:

“The federal government found that 4.5 million children were sexually abused by educators.”

That would be wrong.

Instead:

Shakeshaft extrapolated the 9.6% survey estimate to the contemporary K–12 population and estimated that more than 4.5 million students could experience some form of educator sexual misconduct during their school careers.

That’s accurate.

6. Shakeshaft herself documented the limitations

This actually strengthens our treatment of the study.

She identified several limitations:

students were reporting retrospectively across their school careers;

only grades 8–11 were sampled;

earlier experiences might have been forgotten;

the educator-specific analysis was limited;

and only unwanted behavior was included. (U.S. Department of Education)

But there’s another limitation that we should explicitly add:

Extrapolating grades 8–11 to K–12 is substantially less secure than generalizing the survey to grades 8–11.

The survey design supports the latter.

The K–12 extrapolation requires an additional assumption.

That distinction needs to appear prominently in our report.

7. Here’s something fascinating: other studies produced wildly different estimates

The 2004 review found U.S. estimates ranging from:

3.7% to 50.3%.

That huge spread resulted from differences in sampling, definitions and methodology.

Shakeshaft therefore concluded that the 9.6% AAUW-based estimate was the most accurate available at that time because of its stronger sampling and survey design. (U.S. Department of Education)

That is an important lesson for our entire project:

Methodology matters more than the biggest number.

We’re not going to headline 50.3% because it sounds dramatic.

We’re going to use the strongest evidence.

8. And look what the report said in 2004

This may eventually become one of our most consequential findings.

Shakeshaft wrote that there was:

no national U.S. incidence or prevalence study whose primary purpose was educator sexual abuse.

The available U.S. prevalence evidence consisted of seven studies using six datasets that approached the issue from ancillary or regional perspectives. (U.S. Department of Education)

Remember that.

Because when we reach 2025–2026, we’re going to ask:

Twenty-two years later, has that fundamental national data gap actually been solved?

If the answer is no, that is potentially a major policy finding.

9. Male victims — our Connecticut story has historical context

Remember the New Haven Register article you brought me about female teachers and young male victims?

Shakeshaft was discussing this twelve years earlier.

Student surveys showed a much smaller sex disparity among targets than official-report studies did.

In the AAUW/Shakeshaft analysis:

56% of targets were female.

44% were male.

But studies relying on formal reports produced much larger female-majority distributions.

Shakeshaft explicitly suggested this discrepancy might mean misconduct against male students was less likely to be reported. (U.S. Department of Education)

That deserves its own later subsection:

The Invisible Victim Problem: Boys, Reporting and Cultural Stereotypes

And there was another striking discrepancy.

Student survey reports attributed misconduct to:

57.2% male educators

and

42.8% female educators.

Some official/investigative datasets, however, showed overwhelmingly male perpetrators—sometimes around 80–96%. Shakeshaft discussed the possibility that sexual conduct by women toward boys might be particularly underreported because boys may be culturally conditioned to interpret sexual attention from women as flattering rather than victimization. (U.S. Department of Education)

Again: hypothesis, not established fact.

We’ll label it accordingly.

But it gives historical context to the Connecticut reporting you found.

10. Then we get into grooming

This section is sobering.

The report describes the power imbalance inherent in schools and discusses patterns whereby offending educators may isolate students, create complicity, manipulate them, or select children perceived as vulnerable or unlikely to be believed. (U.S. Department of Education)

And Shakeshaft specifically described much misconduct as occurring within:

“the much subtler framework of grooming and enticement.” (U.S. Department of Education)

This was 2004.

That matters because modern child-protection conversations sometimes make “grooming” sound like newly developed terminology.

It isn’t.

MASTER EVIDENCE TABLE — ENTRY 001

I think this is the form we should preserve:

Field	Finding
Source	Shakeshaft, Charol. Educator Sexual Misconduct: A Synthesis of Existing Literature
Year	2004
Authority	Prepared under contract for U.S. Department of Education pursuant to congressional mandate
Evidence type	Federal literature synthesis + secondary analysis
Primary prevalence dataset	2000 AAUW/Harris Interactive Hostile Hallways
Population sampled	Public-school students, grades 8–11
N	2,065
Nationally representative?	Yes for public-school grades 8–11 according to source methodology
Primary result	9.6% reported unwanted educator sexual misconduct
Contact	6.7%
Non-contact	8.7%
K–12 extrapolation	>4.5 million students
Extrapolation status	Estimate based on assumption—not directly measured
Definition	Broad sexual-misconduct umbrella encompassing verbal, visual and physical conduct
Major strength	Strong national sampling methodology for grades 8–11
Major weakness	Educator misconduct wasn’t original survey’s primary research purpose; K–12 extrapolation extends beyond sampled grades
Our evidence rating	FOUNDATIONAL — with clearly stated qualifications

And I think we have our first permanent editorial rule:

The 9.6% finding is defensible. The 4.5-million figure is an extrapolation. Neither figure means 9.6% of children were raped or subjected to contact sexual abuse.

That sentence alone will protect the credibility of the eventual report from a great deal of sloppy argument.

Next up, I want to move forward chronologically to the Associated Press 2007 investigation—because that gives us an entirely different kind of evidence: not anonymous student reports, but thousands of actual educators who entered state disciplinary systems.

Survey prevalence → documented disciplinary cases.

Now we’re beginning to triangulate the problem from independent directions.

Next…

Onward. This time we move from survey-estimated prevalence to something quite different: educators who actually reached state disciplinary systems.

FOUNDATION STUDY #2

Associated Press Nationwide Investigation, 2007

Sex Abuse a Shadow Over U.S. Schools

This is not a peer-reviewed prevalence study. It is a seven-month, nationwide investigative reporting project in which Associated Press reporters sought state educator-licensing disciplinary records from all 50 states and the District of Columbia for 2001–2005. Most reporters had to make formal public-records requests, sometimes repeatedly; ultimately all states except Maine provided most of the requested information. (Education Week)

That makes this evidence fundamentally different from Shakeshaft—and extremely useful.

1. The headline finding

The AP identified:

2,570 educators

whose teaching credentials were revoked, denied, surrendered or otherwise sanctioned from 2001 through 2005 following allegations of sexual misconduct. (Education Week)

Those cases represented approximately one-quarter of all educator-misconduct cases AP identified during that five-year period. (Education Week)

But here’s our first precision rule:

2,570 is not the estimated number of offending educators in the United States during those five years.

It is the number AP was able to identify through state credential-disciplinary systems.

That distinction is enormous.

2. What happened to the licenses?

AP actually identified 2,625 credential actions, because more than 50 educators were disciplined in more than one state.

Among those actions:

1,636 — licenses revoked

440 — licenses surrendered

376 — licenses suspended

108 — licenses denied

with other sanctions comprising the remainder. (Education Week)

That multi-state duplication is itself interesting.

An educator disciplined in one jurisdiction could potentially appear in another state’s licensing system.

That leads directly toward the problem we’ll eventually examine under:

Interstate Information Sharing and “Passing the Trash”

3. How many involved children?

AP identified a young person as the victim in at least:

1,801 cases

“Young person” included students as well as unidentified youths, family members and neighbors.

Within that group, students were specifically identified as victims in at least 1,467 cases. (Daily Herald)

That’s another important distinction.

We should not report:

2,570 educators sexually abused students.

The evidence doesn’t establish that.

Instead:

AP identified 2,570 educators subjected to credential action following allegations of sexual misconduct; students were clearly identified as victims in at least 1,467 cases.

Much stronger.

Because it’s accurate.

4. Contact misconduct

Among the 1,801 cases involving young victims, AP found physical contact in at least:

1,297 cases — 72%

The remaining cases included non-contact conduct such as verbal sexual harassment and other offenses. (Daily Herald)

This is particularly useful beside Shakeshaft.

Remember:

Shakeshaft’s 9.6% encompasses a broad spectrum of contact and non-contact misconduct.

AP’s disciplinary dataset contains a much larger proportion of serious physical-contact cases.

That makes sense.

The more serious the behavior, the more likely it may be to reach:

complaint → investigation → state licensing authority → disciplinary action.

So the two datasets are looking through different windows at the same underlying problem.

5. Criminal convictions

Here’s an especially important number:

At least 1,390 cases resulted in criminal convictions.

That’s approximately 53% of AP’s 2,570 cases. (Daily Herald)

This gives us a particularly strong evidentiary subset.

We therefore have three progressively narrower categories:

2,570 educators subjected to credential action following sexual-misconduct allegations

↓

at least 1,467 cases clearly identifying students as victims

↓

at least 1,390 cases involving criminal convictions

But—we should not assume the 1,390 convictions are all contained within the 1,467 student-victim cases.

The published summary doesn’t establish that intersection.

That is precisely the kind of statistical shortcut we’re going to avoid.

6. Multiple victims

AP found:

At least 446 cases involved educators with multiple victims. (Daily Herald)

This number matters beyond its shock value.

It raises an institutional question:

How long were some offenders able to continue before effective intervention occurred?

That becomes especially important when we later combine AP with the GAO investigation.

The AP reporting describes cases involving prior warning signs, inadequate responses, quiet departures, confidential agreements and employees subsequently obtaining positions elsewhere. (Education Week)

Now we’re beginning to move beyond prevalence into systems failure.

7. Sex of disciplined educators

Nearly:

9 in 10 disciplined educators were male. (Daily Herald)

Interesting—but remember what we discovered in Shakeshaft.

Anonymous student reporting produced a substantially larger proportion of alleged female perpetrators than official disciplinary datasets did.

That creates a research question:

Why?

Possible explanations include genuine differences in offending behavior, differences in severity, differences in reporting, differences in institutional responses, cultural attitudes toward male victims, or combinations of these.

We do not yet know which explanation accounts for how much of the discrepancy.

So we flag it rather than solve it.

8. A major limitation: licensing records miss people

This is critical.

AP reported that the overwhelming majority of cases it examined involved public-school teachers. Private-school teachers appeared much less frequently because many were not required to hold state teaching credentials and private disciplinary actions were often not publicly accessible. (Education Week)

But there is another limitation inherent in the method.

A licensing database is good at finding:

licensed educators who were reported and disciplined.

It is much worse at finding:

misconduct never reported;

reported conduct never forwarded to licensing authorities;

cases investigated without discipline;

employees who resigned before formal proceedings;

school employees who didn’t require educator licenses;

and potentially misconduct concealed through private employment agreements.

So once again:

AP’s 2,570 is a floor within a particular administrative system—not a prevalence estimate.

9. AP found enormous differences among states

This is potentially very important.

AP found substantial variation among states in the number of educators disciplined—even after accounting for population differences. The investigation connected that variation to differences in state laws, reporting requirements, background checks and disciplinary practices. (Education Week)

That means:

A state reporting fewer disciplined educators does not necessarily have less educator sexual misconduct.

It could have:

better prevention,

or fewer incidents,

or weaker detection, reporting and disciplinary systems.

Those possibilities cannot be distinguished merely from the raw disciplinary count.

This same problem occurs constantly in child-protection statistics.

Recorded incidence ≠ underlying incidence.

10. And then AP encountered secrecy

Here’s where our future transparency section begins.

Maine was the only state that refused to disclose its educator sexual-misconduct cases because state law kept offending teachers’ names secret; AP independently located three publicly reported Maine cases. (Education Week)

California presented another problem.

AP reported that more than 300 California educators had licenses revoked or suspended for sex-related offenses during 2001–2005, but state confidentiality rules prevented the public from readily determining this from enforcement records—even in some situations involving educators who had gone to prison and registered as sex offenders. (Education Week)

Now THAT deserves a flag in our evidence file:

Transparency can directly affect apparent prevalence.

If records are inaccessible, cases don’t disappear.

They disappear from the dataset.

11. The most important conceptual contribution of the AP investigation

Shakeshaft asks:

What do students report experiencing?

AP asks:

What appears in official state disciplinary records?

Those are radically different measurements.

And look at the conceptual pipeline emerging:

THE ATTRITION PYRAMID

Actual misconduct

↓

Misconduct recognized by student/others

↓

Disclosed

↓

Reported to institution

↓

Investigated

↓

Reported to state licensing authority

↓

Credential action

↓

Criminal investigation

↓

Prosecution

↓

Conviction

Every step can lose cases.

Therefore:

The number of convictions can never, by itself, tell us how much misconduct occurred.

Likewise:

The number of allegations can never, by itself, tell us how much misconduct was substantiated.

Both statements belong in our report.

MASTER EVIDENCE TABLE — ENTRY 002

Field	Finding
Source	Associated Press, Sex Abuse a Shadow Over U.S. Schools / A Lingering Shame
Year published	2007
Period examined	2001–2005
Evidence type	Nationwide investigative analysis of state disciplinary records
Jurisdictions approached	All 50 states + District of Columbia
Educators identified	2,570
Credential actions	2,625
Revocations	1,636
Surrenders	440
Suspensions	376
License denials	108
Students clearly identified as victims	≥1,467 cases
Young victims identified	≥1,801 cases
Physical contact among youth-victim cases	≥1,297 / 72%
Criminal convictions	≥1,390 / 53% of identified cases
Multiple-victim cases	≥446
Male disciplined educators	Nearly 90%
Primary strength	Nationwide records collection from state licensing authorities
Primary weakness	Captures only cases entering credential-disciplinary systems; not prevalence
Our evidence rating	MAJOR NATIONAL DOCUMENTARY EVIDENCE

Primary AP reporting is preserved through Education Week’s publication of the Associated Press investigation. Its methodology is separately documented in How the AP project unfolded.

Now put 2004 and 2007 beside each other

This is where our report begins becoming powerful.

2004 — U.S. Department of Education

Student survey data suggest educator sexual misconduct is far more common than official cases would imply.

2007 — Associated Press

State government records document thousands of actual disciplinary cases, including more than a thousand criminal convictions, while simultaneously revealing inconsistent reporting, confidentiality barriers and substantial differences among state systems.

Two completely different methods.

Two completely different datasets.

Both pointing toward a significant national problem.

And neither can tell us the true annual national incidence.

And NOW we reach a pivotal investigation

Our next major evidence source is:

U.S. GOVERNMENT ACCOUNTABILITY OFFICE — 2010

And this one changes the question.

It isn’t principally:

“How many?”

It is:

“How did known or suspected offenders remain in—or return to—positions where they could reach children?”

GAO examined actual cases involving convicted offenders and discovered failures involving hiring, references, background checks, reporting and school-to-school movement.

This is where “passing the trash” stops being merely a disturbing phrase and becomes a documented institutional child-safety problem.

That’s the next section…

Yes. Now we reach one of the strongest pieces of government evidence about institutional failure in the entire project.

FOUNDATION STUDY #3

U.S. Government Accountability Office — 2010

K–12 Education: Selected Cases of Public and Private Schools That Hired or Retained Individuals with Histories of Sexual Misconduct

This is GAO-11-200, published December 8, 2010.

U.S. GAO — Full report and supporting materials

This report matters because GAO was no longer asking primarily:

How common is educator sexual misconduct?

It was investigating:

How did people with histories of sexual misconduct get hired or remain employed in schools—and what failures allowed that to happen?

That is an entirely different question.

And GAO found some extraordinary things.

1. What GAO actually investigated

GAO was asked to examine circumstances in which K–12 schools hired or retained people with histories of sexual misconduct and determine what factors contributed to those employment decisions.

This included both public and private schools.

GAO compared 2007–2009 employment databases from 19 states and the District of Columbia against information in the National Sex Offender Registry. It also searched public records from 2000–2010 for school-employee sexual-misconduct cases that ultimately resulted in criminal convictions.

From that work, investigators selected:

15 cases in 11 states

for detailed investigation.

They examined court documents and personnel records and, where possible, interviewed school officials and law-enforcement personnel. (Government Accountability Office)

VERY IMPORTANT

These 15 cases were case studies.

They were not randomly sampled.

Therefore:

We cannot use 15 cases to estimate national prevalence.

But that’s not what makes this report important.

The value lies in what investigators discovered about how institutional safeguards failed.

2. Eleven of the 15 had previously targeted children

This is one of the strongest findings:

At least 11 of the 15 cases involved offenders who had previously targeted children.

And then:

In at least 6 cases, offenders used their new positions in schools to abuse additional children. (Government Accountability Office)

That distinction is crucial.

These weren’t merely people whose pasts theoretically presented some statistical risk.

In at least six cases:

prior history

↓

subsequent access to children through another school position

↓

additional children abused

That is documented institutional child-safety failure.

3. GAO identified FOUR recurring failures

This should become a major box in our eventual report.

FAILURE #1 — Allowing resignation instead of disciplinary action

GAO found cases in which officials allowed educators who had engaged in sexual misconduct toward students to resign rather than face disciplinary action.

Worse:

Some subsequently received positive references.

(Government Accountability Office)

That is essentially the mechanism later widely described as:

PASSING THE TRASH

An employee leaves.

The institution avoids a difficult disciplinary process.

The employee’s record doesn’t adequately communicate what occurred.

Another school hires the employee.

Children are placed at risk again.

4. FAILURE #2 — No criminal-history check

GAO found schools that simply didn’t conduct pre-employment criminal-history checks. (Government Accountability Office)

But this leads to something even more important.

5. FAILURE #3 — Background checks that weren’t good enough

Some schools did conduct background checks.

And still failed.

GAO identified problems because checks weren’t necessarily:

national,

fingerprint-based,

recurring. (Government Accountability Office)

That’s an important policy distinction.

It isn’t enough for legislators to say:

“Schools conduct background checks.”

The meaningful questions are:

What databases?

Which jurisdictions?

Fingerprint or name-based?

Before employment only?

Repeated periodically?

Are educator-disciplinary databases checked?

Does the hiring school obtain information from previous employers?

That’s much more sophisticated child-safety analysis.

6. FAILURE #4 — They sometimes had the information and didn’t act on it

This one is almost difficult to believe.

GAO found schools that failed to investigate troubling criminal-history information disclosed on employment applications. (Government Accountability Office)

Meaning the problem wasn’t always:

“Nobody could possibly have known.”

Sometimes information was available.

The system simply failed to use it.

That distinction will matter enormously in our final report:

INFORMATION FAILURE

versus

ACTION FAILURE

A child-protection system can fail because information isn’t collected.

It can also fail because information is collected and then ignored.

7. Case Study: Ohio

This example practically defines passing the trash.

An Ohio teacher was forced to resign after inappropriate conduct involving female students.

And yet his superintendent subsequently wrote him a recommendation describing him as an:

“outstanding teacher.”

He obtained employment in a neighboring school district.

There, he was ultimately convicted of sexual battery against a sixth-grade girl. (Government Accountability Office)

This belongs in our report.

Not because we want sensational individual stories.

Because it demonstrates the mechanism.

Prior warning.

↓

Resignation.

↓

Positive reference.

↓

New school.

↓

New access to children.

↓

Another victim.

That is precisely the institutional phenomenon policymakers need to understand.

8. Case Study: Arizona

This one demonstrates a different failure.

A school was rushing to fill a position.

The applicant himself disclosed that he had committed a dangerous crime against a child.

The school hired him without completing the criminal-history check first.

He was subsequently convicted of sexual contact with a young female student.

Investigators also found videos of nude underage girls in his possession. (Government Accountability Office)

Again:

The information wasn’t necessarily unknowable.

The protective process wasn’t followed.

9. Case Study: Louisiana

GAO found a teacher who was already a registered sex offender and whose Texas teaching certificate had been revoked.

He nevertheless obtained employment at several Louisiana schools without receiving the appropriate criminal-history check.

He was later accused of engaging in sexual conversations with a student. (Government Accountability Office)

Now we have another important category:

INTERSTATE INFORMATION FAILURE

What did Texas know?

What did Louisiana know?

Could Louisiana have discovered what Texas knew?

Was anyone required to ask?

Was anyone required to answer?

Those questions eventually lead directly into federal “passing the trash” legislation.

10. Case Study: California

This one is extraordinary for another reason.

A school employee was convicted in 2000 of molesting a minor.

According to GAO:

The school knew about the conviction and did not fire him.

He remained employed at the California public school until 2010.

After GAO referred the matter to the California Attorney General, officials placed him on administrative leave; he subsequently resigned. (Government Accountability Office)

That is almost a textbook example of why we need the category:

Known Information — No Effective Action

11. And then GAO found a legal patchwork

At the time, GAO found:

No federal law regulating the employment of sex offenders in public or private schools.

State laws varied dramatically.

Some states required national fingerprint-based background checks.

Others didn’t.

States also differed regarding whether convictions required:

termination,

license revocation,

or refusal to hire. (Government Accountability Office)

So in 2010, a child’s protection could depend substantially upon:

which side of a state line the school happened to be located on.

That belongs in our policy analysis.

12. Something happened AFTER this report

This is where our government-response timeline starts becoming fascinating.

The GAO documented the problem in 2010.

Congress eventually addressed “passing the trash” explicitly in the Every Student Succeeds Act of 2015, requiring states receiving federal education funds to have laws, regulations or policies prohibiting school employers from assisting employees in obtaining new education employment when the employer knows or has probable cause to believe the employee engaged in sexual misconduct with a minor or student, subject to specified exceptions.

That means we will eventually be able to ask:

Did the federal intervention work?

And fortunately, later research investigated exactly that.

We’ll get there.

But now I found something we MUST add to our timeline

Our previous outline jumped from GAO 2010 toward later research.

There is another major GAO report:

GAO — 2014

Child Welfare: Federal Agencies Can Better Support State Efforts to Prevent and Respond to Sexual Abuse by School Personnel

And THIS is enormously important.

GAO surveyed educational agencies in all 50 states and D.C. (Government Accountability Office)

They found:

46 states + D.C. required background checks for at least some public-school applicants.

But the methods varied considerably.

42 states had professional standards/codes of conduct.

Only 22 included information about appropriate employee/student boundaries.

And only:

18 states

reported requiring school districts to provide training addressing sexual abuse or misconduct. (Government Accountability Office)

That’s remarkable.

13. Reporting requirements had problems too

GAO found:

46 states had laws requiring school personnel to report child sexual abuse.

43 states established penalties for failure to report. (Government Accountability Office)

Sounds pretty good.

Until investigators looked at what actually happened.

Three of six school districts GAO visited had policies instructing employees to report suspected abuse/misconduct to school administrators.

Local investigators told GAO that such policies could cause confusion by implying that reporting internally satisfied the employee’s obligation.

And GAO documented a case where administrators pleaded guilty to failing to report suspected sexual abuse by a teacher.

That teacher was subsequently convicted on:

10 counts of abuse. (Government Accountability Office)

Again:

policy exists

does not necessarily equal

child protected.

14. And here comes one of the most important sentences in our entire project

Remember our question:

Twenty-two years after Shakeshaft, do we actually know the national prevalence?

Look at what GAO concluded in 2014:

“the prevalence of this type of abuse is not known.” (Government Accountability Office)

And GAO explained why.

Several federal agencies collected related information, but:

none systematically identified the extent of sexual abuse by school personnel. (Government Accountability Office)

STOP.

Put a giant research flag beside that.

Because we now have:

2004 — Department of Education

No adequate national prevalence study specifically designed around educator sexual misconduct.

2014 — Government Accountability Office

Prevalence still unknown.

Federal agencies collect related information, but none systematically measures its extent.

That is now a decade of documented federal data failure.

And we’re going to follow that thread forward.

15. Federal coordination was also weak

GAO found that Education, HHS and Justice all had programs that could contribute to prevention and response.

But:

More than 30 states surveyed weren’t aware of available federal resources.

And:

No single federal agency was leading the effort.

GAO characterized federal coordination as limited. (Government Accountability Office)

That is not an advocacy organization’s allegation.

That’s the United States Government Accountability Office.

MASTER EVIDENCE TABLE — ENTRY 003

Field	Finding
Source	U.S. Government Accountability Office
Report	GAO-11-200
Year	2010
Evidence type	Federal investigative case study
Cases investigated	15
States represented	11
Previously targeted children	≥11 of 15
Used subsequent school position to abuse additional children	≥6 of 15
Major failures	Resignation instead of discipline; positive references; missing/inadequate background checks; failure to investigate disclosed warning information
National prevalence study?	NO
Can percentages from 15 cases be generalized nationally?	NO
Primary value	Documents mechanisms allowing known/high-risk individuals continued access to children
Evidence rating	MAJOR FEDERAL INSTITUTIONAL-FAILURE EVIDENCE

(Government Accountability Office)

MASTER EVIDENCE TABLE — ENTRY 004

Field	Finding
Source	U.S. Government Accountability Office
Report	GAO-14-42
Year	2014
Scope	50 states + D.C.; four states/six districts examined more closely
Background-check requirements	46 states + D.C.
Professional standards/codes	42 states
Codes addressing boundaries	22 states
States reporting required district prevention training	18
States requiring personnel to report child sexual abuse	46
States establishing penalties for failure to report	43
National prevalence known?	NO
Federal system systematically measured prevalence?	NO
Federal coordination	GAO found it limited; no single agency led the effort
Evidence rating	MAJOR FEDERAL SYSTEMS EVIDENCE

(Government Accountability Office)

Look at our chronology now

2004 — Department of Education:
Substantial student-reported misconduct; national measurement inadequate.

↓

2007 — Associated Press:
Thousands of educators appear in disciplinary records; major differences and gaps among state systems.

↓

2010 — GAO:
Schools sometimes hire or retain people despite histories of sexual misconduct; some subsequently abuse additional children.

↓

2014 — GAO:
State prevention/reporting systems remain inconsistent—and the federal government still cannot tell us the national prevalence.

↓

2015 — Congress:
ESSA specifically addresses assisting employees accused of sexual misconduct in obtaining new school employment.

And then comes the obvious question:

WHAT HAPPENED AFTER CONGRESS ACTED?

Did all 50 states implement effective anti-passing-the-trash protections?

Did background checks improve?

Did states begin sharing information?

Did national measurement improve?

Did schools stop quietly allowing accused employees to move elsewhere?

That brings us to the post-ESSA compliance research—and I think it is going to become another major chapter of What the Data Actually Show.

Because the answer was not simply:

“Congress passed a law; problem solved.”

Next…

Excellent. Now we can test what happened after Congress acted. This is especially valuable because we have both an independent federally funded study shortly after ESSA and a much more extensive U.S. Department of Education study several years later.

FOUNDATION STUDY #4

ESSA and “Passing the Trash”: Did Federal Law Fix the Problem?

In December 2015, Congress enacted the Every Student Succeeds Act (ESSA). Section 8546 addressed something the earlier evidence had already exposed: schools facilitating—or failing to prevent—the movement of employees suspected of sexual misconduct into other education jobs.

The provision requires states receiving ESEA funds to have laws, regulations or policies prohibiting school employers from assisting an employee in obtaining new school employment when they know, or have probable cause to believe, that the employee engaged in sexual misconduct with a student or minor in violation of law, subject to limited exceptions.

This is important conceptually:

Congress did not require a criminal conviction before the protection could apply.

The statutory threshold includes circumstances involving probable cause to believe sexual misconduct occurred, with procedural exceptions.

That’s because waiting for a criminal conviction would leave a very large gap in a child-safety system.

1. Then researchers asked the obvious question

Were the states actually complying?

Billie-Jo Grant, Stephanie Wilkerson and Molly Henschel conducted a federally funded study examining state implementation.

The research was supported by the National Institute of Justice, U.S. Department of Justice, and subsequently appeared in the peer-reviewed Journal of Child Sexual Abuse. (PubMed)

Researchers contacted state departments of education and investigated whether states had developed law or policy meeting the new ESSA requirement.

The results were startling.

Only FOUR states were found to have fully complied.

They were:

Connecticut
Oregon
Pennsylvania
Washington

Several additional states were working on legislation or policy.

But:

39 states

had no plans to create relevant legislation or policy at the time of the study. (PubMed)

And WHY they weren’t doing it is especially interesting.

2. Some states didn’t know they needed to do anything

Researchers found two principal explanations among the 39 states:

Some state officials were unaware of the federal provision.

Others believed their existing laws already satisfied ESSA when the researchers concluded that they did not. (PubMed)

In fact, researchers reported that a few states began addressing the issue because the researchers asked them about it. (National Institute of Justice)

Think about the chronology:

2010: GAO documents institutional failures allowing offenders continued access to children.

↓

2014: GAO finds inconsistent protections and inadequate national systems.

↓

2015: Congress legislates.

↓

2016–17: Researchers contact states.

↓

Only four states fully satisfy the researchers’ interpretation of the federal requirement.

That is an important policy story.

3. But we need a methodological caution here

This study is valuable, but it wasn’t a federal compliance audit.

Researchers collected information from state education officials through telephone and email contacts, supplemented by other sources. Seventeen states responded by phone and 18 by email; information for nine states was supplied by S.E.S.A.M.E., while six states either did not respond or pointed researchers to information that the researchers concluded did not satisfy the requirement. (Taylor & Francis Online)

Therefore our report should say:

Grant, Wilkerson and Henschel concluded that four states had fully complied with the ESSA provision at the time of their study.

Not:

“The federal government found only four states were legally compliant.”

It didn’t.

That distinction matters.

4. Then the U.S. Department of Education conducted a much larger study

And THIS gives us a fascinating later comparison.

In 2022, the Department published:

Study of State Policies to Prohibit Aiding and Abetting Sexual Misconduct in Schools

The researchers systematically reviewed laws, regulations, codes of ethics and policies across:

all 50 states and the District of Columbia

and conducted interviews with state education agency personnel.

The documentary findings were verified by 48 of the 51 state education agencies. (U.S. Department of Education)

That’s a substantial evidence source.

5. By 2022, things had improved

The Department found that all 50 states and D.C. required prospective school employers to conduct criminal-background checks.

And:

46 of 51 required fingerprinting of applicants.

(U.S. Department of Education)

Moreover:

35 states

had enacted at least one additional provision intended to prohibit aiding and abetting the employment of individuals who had engaged—or were alleged to have engaged—in sexual misconduct with a student or minor. (U.S. Department of Education)

That is real progress.

Our report should say so.

We’re not trying to prove that government did nothing.

We’re trying to find out what actually happened.

6. But look what happens when we go beyond criminal background checks

This is extremely important.

A criminal-background check can only reveal something that has entered the criminal system.

Remember our attrition pyramid?

Misconduct

↓

disclosure

↓

report

↓

investigation

↓

possible discipline

↓

possible criminal investigation

↓

possible prosecution

↓

possible conviction.

So checking criminal history is useful—but it cannot detect every employee who has previously engaged in concerning conduct.

And the Department’s own numbers illustrate the problem.

Only:

27 states

required prospective employers to check some combination of an applicant’s:

employment history,
certification status,
employment eligibility, and/or
disciplinary status.

Among those:

19 states required information to be requested from current/former employers.

14 states required checking employment or certification eligibility within or across states.

And only:

11 states

required applicants to disclose information regarding investigations or disciplinary actions related to sexual abuse or misconduct. (U.S. Department of Education)

That deserves attention.

7. This gives us a very important distinction

CRIMINAL BACKGROUND CHECK

asks:

Has this person entered the criminal-justice system?

EMPLOYMENT / DISCIPLINARY HISTORY

asks:

Has something concerning happened in this person’s previous professional employment?

Those are not interchangeable.

Someone can have:

no criminal conviction

and simultaneously have:

a substantiated employment investigation,

professional discipline,

license action,

a resignation during an investigation.

That is exactly why “passing the trash” exists as a policy problem in the first place.

8. Then comes another major finding

Of the state education agencies studied, only:

7 SEAs

reported documenting complaints and/or incidents of sexual misconduct from districts.

33 said they did not.

Another 8 did not know whether their SEA did so. (U.S. Department of Education)

Now stop there.

This connects directly to our larger project.

We’ve been asking:

Why is national prevalence so difficult to establish?

Here is part of the answer.

If state education agencies themselves frequently aren’t collecting district-level sexual-misconduct complaints/incidents, there cannot simply be a clean national database created by adding the states together.

Some information instead resides with professional-standards boards, licensing agencies or other government bodies. (U.S. Department of Education)

In other words:

The data are institutionally fragmented.

That’s going into our report.

9. Even having a law doesn’t tell us whether districts obey it

This may be one of my favorite findings—not because it’s good news, but because it demonstrates precisely the analytical problem we’re uncovering.

State officials told researchers that they didn’t necessarily have mechanisms to monitor whether individual districts actually complied.

One official essentially described the problem as:

“we don’t know what we don’t know.”

The state could require districts to contact previous employers, for example, without possessing a system capable of verifying that every district actually did it for every applicant. (U.S. Department of Education)

THAT gives us another layer:

LAW EXISTS

does not necessarily mean

POLICY EXISTS

does not necessarily mean

DISTRICT IMPLEMENTS POLICY

does not necessarily mean

COMPLIANCE IS MONITORED

does not necessarily mean

CHILD IS PROTECTED.

That sequence belongs in the finished report.

10. States reported barriers

State officials described difficulties including:

privacy laws restricting disclosure of personnel records;

stakeholder or legislative opposition;

complex statutory drafting;

continuing communication with districts;

lack of mechanisms for monitoring local compliance;

and

resources required for background and employment-history checks. (U.S. Department of Education)

The process of developing these laws was described by some officials as “painstaking” and “intensive.” (U.S. Department of Education)

And states wanted more help from Washington.

19 states wanted examples of effective state/district policies.

12 states wanted clearer federal guidance.

Several called for additional federal assistance. (U.S. Department of Education)

So again, the evidence doesn’t support a simplistic narrative of:

“Schools don’t care.”

The reality is more complicated.

There are legal conflicts, privacy issues, labor considerations, decentralized governance, information silos, administrative capacity problems—and sometimes outright institutional failure.

We document all of it.

MASTER EVIDENCE TABLE — ENTRY 005

Field	Finding
Study	Grant, Wilkerson & Henschel, Passing the Trash
Published	2018 online / 2019 journal issue
Funding	National Institute of Justice, U.S. DOJ
Evidence type	State-policy implementation research
Federal provision examined	ESSA §8546
States researchers found fully compliant	4
States with no plans for additional legislation/policy	39
Important qualification	Researcher assessment—not federal legal compliance determination
Primary significance	Demonstrated slow/inconsistent early state implementation after federal legislation
Evidence rating	IMPORTANT PEER-REVIEWED POLICY EVIDENCE

(PubMed)

MASTER EVIDENCE TABLE — ENTRY 006

Field	Finding
Source	U.S. Department of Education
Report	Study of State Policies to Prohibit Aiding and Abetting Sexual Misconduct in Schools
Published	2022
Jurisdictions examined	50 states + D.C.
Criminal background checks	51/51
Fingerprinting	46/51
States with ≥1 additional anti-aiding/abetting provision	35
Employment/certification/disciplinary-history requirements	27
Require information from former/current employers	19
Cross-state employment/certification eligibility checks	14
Applicant disclosure of misconduct investigations/discipline	11
SEA documenting district misconduct complaints/incidents	7
SEA not documenting them	33
SEA respondents didn’t know	8
Primary strength	Systematic federal study covering all states + D.C.
Primary limitation	Primarily measures existence/implementation of laws and policies—not prevalence or effectiveness at preventing abuse
Evidence rating	MAJOR FEDERAL POLICY/SYSTEMS EVIDENCE

(U.S. Department of Education)

And now something bigger is emerging

Look at what we have established without making a single speculative claim:

2004: National prevalence measurement inadequate.

2007: Thousands of disciplinary cases documented.

2010: GAO documents schools hiring/retaining people with known histories and subsequent victims.

2014: GAO says national prevalence remains unknown and prevention systems vary substantially.

2015: Congress acts.

2016–17: Researchers find very limited early state implementation of the new federal requirement.

2022: Federal research finds meaningful improvement—but substantial differences remain in employment-history checking, information sharing, state data collection and compliance monitoring.

This suggests something deeper than simply “bad teachers.”

We are looking at an information-system problem.

Who knows what?

Who must report it?

Who records it?

Who investigates it?

Who can access it?

Who must disclose it?

Who checks it before hiring?

What happens when an employee crosses district lines?

What happens when an employee crosses state lines?

What happens when allegations exist but no criminal conviction does?

What happens when an employee resigns during an investigation?

And perhaps most importantly:

Who is responsible for connecting all of those pieces?

That may eventually become one of this report’s major policy contributions.

Because background checks alone cannot solve a problem when the relevant information never entered the criminal database.

And now we are ready for another turn in the evidence.

The next major section should move away from institutional-policy research and return to students themselves:

Jeglic et al. — contemporary prevalence research

This gives us an opportunity to ask an enormously important question:

Almost twenty years after Shakeshaft’s 9.6% finding, did newer research find the problem had substantially disappeared?

We already know the short answer.

No.

But now we’ll examine exactly what the newer study measured—and, just as importantly, what it didn’t measure.

Next…

Absolutely. Now we return to the students themselves, almost twenty years after Shakeshaft.

And this study is important because it gives us something we have been missing: a large, contemporary, multistate sample of recent high-school graduates.

FOUNDATION STUDY #5

Jeglic et al. — 2023

The Nature and Scope of Educator Misconduct in K–12

Published in the peer-reviewed journal Sexual Abuse, this study was conducted by Elizabeth Jeglic and seven colleagues affiliated with institutions including John Jay College of Criminal Justice and George Mason University. It was first published online in May 2022 and appeared in the March 2023 journal issue. (PubMed)

The researchers explicitly began with the problem we’ve been documenting:

Shakeshaft was nearly twenty years old, and there had been relatively little large-scale research since then. (PubMed)

So they asked:

What are recent students actually reporting?

1. The sample was LARGE

The researchers surveyed:

6,632 recent high-school graduates

from four U.S. states about experiences occurring during grades K–12. (PubMed)

That’s substantially larger than the approximately 2,000-student dataset underlying Shakeshaft’s principal national estimate.

But—and this is critical—

It was NOT a nationally representative sample.

It was a large four-state sample.

Therefore we absolutely should not write:

“11.7% of all American students experience educator sexual misconduct.”

The study doesn’t establish that.

What it does establish is:

Among 6,632 recent high-school graduates surveyed across four U.S. states, 11.7% reported experiencing at least one form of educator sexual misconduct during grades K–12. (PubMed)

That is plenty powerful without embellishment.

2. The headline number

The result was:

11.7%

reporting at least one form of educator sexual misconduct. (PubMed)

At first glance, someone might compare:

Shakeshaft 2004 — 9.6%

with

Jeglic 2023 — 11.7%

and conclude:

“Educator sexual misconduct increased.”

NO.

We cannot say that.

Different samples.

Different methodologies.

Different time periods.

Different measures.

Different populations.

One nationally representative grades 8–11 dataset versus one large four-state sample of recent graduates recalling K–12 experiences.

The correct conclusion is:

A large contemporary study conducted nearly two decades after Shakeshaft continued to find substantial self-reported educator sexual misconduct.

That’s defensible.

3. And THIS is why definitions matter so much

Of the entire sample:

11.0% reported sexual comments.

But the rates for other forms of misconduct were each:

less than 1%.

Those included behaviors such as receiving sexual photographs/messages, sexual touching, kissing and sexual intercourse/oral sex. (PubMed)

Jeglic’s own public explanation gives additional detail:

0.6% — shown/given sexual pictures or sent sexual messages

0.9% — sexually touched

0.4% — other sexual activity, including intercourse/oral sex. (Elizabeth L. Jeglic)

This is enormously important for our report.

Someone could accurately write:

“11.7% reported educator sexual misconduct.”

But if they then implied:

“11.7% reported physical sexual abuse,”

they would grossly misrepresent the research.

We’re not going to do that.

4. But don’t make the opposite mistake either

There is another potential distortion:

“Oh. Most of it was just comments.”

That’s not an adequate interpretation either.

Why?

Because the study found that participants reporting educator sexual misconduct also demonstrated significantly greater difficulties in current psychosocial functioning than participants who did not report misconduct. (PubMed)

The study therefore gives us evidence that even though the broad prevalence number is dominated by sexual comments, the reported experiences were associated with measurable differences in later functioning.

We should be cautious about causal language—the study does not establish that educator misconduct necessarily caused every subsequent difficulty.

So:

Associated with — YES.

Proven to have caused — NO.

Another permanent editorial rule.

5. Then comes one of the biggest findings in our project

Remember our attrition pyramid?

The study found that disclosure to authorities was:

4%.

(PubMed)

Now think about what that means methodologically.

Suppose we only study:

police reports,

criminal charges,

teacher-license sanctions,

district investigations,

or court cases.

We’re examining misconduct after it has passed through a disclosure filter.

Jeglic’s findings suggest that, within this sample, disclosure to authorities was extraordinarily uncommon.

That doesn’t tell us exactly what percentage of all U.S. educator misconduct reaches authorities.

But it provides strong empirical evidence that:

Official administrative and criminal records may capture only a subset of student-reported educator sexual misconduct.

THIS is why the AP disciplinary data and the student prevalence surveys don’t conflict.

They’re measuring different stages of the pipeline.

6. Now we can improve our attrition model

I want this eventually illustrated professionally:

THE VISIBILITY GAP

STUDENT EXPERIENCES CONDUCT

↓

Student recognizes conduct as inappropriate

↓

Student tells someone

↓

Student discloses to authority

↓

School records complaint

↓

Investigation occurs

↓

Allegation substantiated

↓

Licensing authority notified

↓

Employment/credential action

↓

Law enforcement referral

↓

Criminal charge

↓

Conviction

And information can disappear at every single stage.

That is one of the central concepts emerging from our evidence.

7. Who were the reported perpetrators?

Among educators reported to have engaged in sexual misconduct:

85% were male.

Academic teachers accounted for:

63%

and coaches/gym teachers:

20%. (PubMed)

Again, that gives us an interesting comparison with Shakeshaft.

Shakeshaft’s older survey analysis produced a considerably larger reported female-perpetrator proportion.

Jeglic produces:

85% male / 15% female.

That does not mean one study must be wrong.

Differences might arise from sampling, question wording, student populations, definitions, time periods, reporting patterns or actual changes.

Flag it.

Don’t explain it away.

8. Who reported experiencing misconduct?

The students reporting educator sexual misconduct were predominantly:

female — 72%.

(PubMed)

Again, however, that means approximately 28% were not female within the reported victim distribution.

That matters when we eventually write the section on boys and male victims.

And we now have multiple sources pointing toward the importance of examining sex differences carefully rather than treating educator misconduct exclusively as a female-student issue.

9. Now we get to grooming

This deserves its own eventual chapter.

Jeglic and colleagues reported grooming-type behaviors including:

29% — special attention

and

12% — gift giving.

(PubMed)

But I want us to be very careful here.

Giving a student special attention or a gift is not inherently evidence of sexual grooming.

Teachers routinely give children attention, encouragement and sometimes small gifts for completely innocent reasons.

Grooming is understood through context, pattern, intent, boundary violations and progression, not by taking an isolated benign behavior and labeling it predatory.

That distinction needs to be explicit in our report.

Otherwise prevention efforts can become both unfair to educators and ineffective at identifying actual grooming patterns.

10. And this is where modern research is becoming more sophisticated

Jeglic isn’t simply studying:

“Did sexual contact occur?”

The research asks about a behavioral continuum:

sexualized comments

↓

boundary-testing / grooming behaviors

↓

sexual communications/images

↓

physical sexual contact

↓

sexual acts

That’s potentially useful for prevention.

If institutions only respond once contact sexual abuse has occurred, then by definition the child-protection system is operating very late.

The harder question becomes:

What earlier behaviors reliably distinguish predatory grooming from normal educator/student interaction?

That is precisely the kind of question modern grooming research is attempting to answer.

And Jeglic is particularly relevant here because sexual grooming and sexual-violence prevention are major areas of her research at John Jay College. (John Jay College)

We’ll return to that separately.

11. Now compare 2004 and 2023 carefully

This is going to become a table in our final publication.

	Shakeshaft 2004	Jeglic et al. 2023
Sample	~2,000 students	6,632 recent graduates
Geographic design	National public-school sample	Four states
Period recalled	School career	K–12
Broad misconduct finding	9.6%	11.7%
Includes non-contact behavior	Yes	Yes
Includes contact behavior	Yes	Yes
National prevalence estimate?	Stronger basis for sampled grades; K–12 extrapolation required assumption	No
Major value	Foundational national evidence	Large contemporary corroborating evidence
Can directly compare percentages as trend?	NO	NO

The conclusion isn’t:

“It rose from 9.6% to 11.7%.”

The conclusion is:

Two large studies conducted roughly two decades apart, using different samples and methods, both found nontrivial levels of self-reported educator sexual misconduct. (PubMed)

That is considerably harder to attack.

MASTER EVIDENCE TABLE — ENTRY 007

Field	Finding
Study	Jeglic et al., The Nature and Scope of Educator Misconduct in K–12
Journal	Sexual Abuse
Published	Online 2022; journal issue 2023
Evidence type	Peer-reviewed empirical research
Sample	6,632 recent high-school graduates
Geographic scope	Four U.S. states
Nationally representative?	NO
Any educator sexual misconduct	11.7%
Sexual comments	11.0%
Sexual pictures/messages	0.6%
Sexual touching	0.9%
Other sexual activity including intercourse/oral sex	0.4%
Disclosure to authorities	4%
Academic teachers among reported perpetrators	63%
Coaches/gym teachers	20%
Male reported perpetrators	85%
Female students among those reporting misconduct	72%
Special attention grooming behavior reported	29%
Gift giving	12%
Major strength	Very large, contemporary multistate sample
Major limitation	Four-state retrospective sample; cannot be converted directly into national prevalence
Evidence rating	MAJOR CONTEMPORARY PEER-REVIEWED EVIDENCE

(PubMed)

And then—almost immediately—another research team found something else

This is where our triangulation becomes even more interesting.

Billie-Jo Grant, Jeffrey Haverland and Jessica Kalbfleisch conducted a different retrospective study of 511 young adults.

Their focus wasn’t precisely the same as Jeglic’s.

They were studying:

Title IX policy implementation AND sexual harassment prevalence in K–12 schools.

Their findings:

17.4%

personally reported one or more incidents of staff-to-student sexual harassment.

And:

50.1%

said they either personally experienced sexual harassment by a school employee or knew someone who had. (Sage Journals)

But here’s what really caught my attention:

94% could not identify their school’s Title IX compliance officer.

And:

70% did not know how to file a grievance.

(Sage Journals)

Now we’re connecting two halves of our investigation:

Student experiences

and

institutional reporting systems.

Because having a Title IX policy sitting in a binder somewhere doesn’t accomplish very much if the students supposedly protected by it don’t know whom to contact or how to use it.

Even more interestingly, Grant’s study reported statistically significantly higher harassment rates among respondents describing districts without sexual-harassment policies/materials than among those reporting such policies/materials. (Sage Journals)

That needs careful examination before we draw causal conclusions.

But it gives us our next foundation study:

Grant et al. — Title IX, sexual harassment and the accessibility of reporting systems

And THAT may help us answer an enormously important child-protection question:

It isn’t enough to ask whether a school has a reporting procedure. Can a child actually find it, understand it and safely use it?

That’s next.

And onward we go. This one adds an important new dimension: having a protection on paper is not the same thing as students knowing how to use it.

FOUNDATION STUDY #6

Grant, Haverland & Kalbfleisch

Title IX Policy Implementation and Sexual Harassment Prevalence in K–12 Schools

The study was first published online in April 2023 and appeared in the 2024 volume of the peer-reviewed journal Educational Policy. The researchers surveyed 511 young adults using a 34-item anonymous online questionnaire about their K–12 experiences, including sexual harassment, school policies, reporting procedures and Title IX. (Sage Journals)

A useful detail for our Master Evidence Table: the respondents represented 35 states. That makes it geographically broad—but it still was not a nationally representative probability sample, so the 17.4% finding cannot legitimately be converted into “17.4% of all American students.” (NSVRC)

And this study actually gives us three different numbers that answer three different questions.

1. Personally experienced staff-to-student sexual harassment

The major finding was:

17.4% — 89 of 511

reported personally experiencing one or more incidents of staff-to-student sexual harassment during K–12. (Sage Journals)

Notice the terminology.

The study says:

sexual harassment

not necessarily:

contact sexual abuse.

That distinction is essential.

We should therefore never write:

“17.4% of students were sexually abused by educators.”

Nor should we write:

“17.4% of American students experience educator sexual harassment.”

Neither statement is supported by this study.

Our wording:

In a retrospective survey of 511 young adults from 35 states, 17.4% reported personally experiencing one or more incidents of staff-to-student sexual harassment during their K–12 education. (Sage Journals)

That is the finding.

2. Then comes the 50.1% number

This is another statistic that could easily be abused.

50.1%

reported that they either personally experienced sexual harassment by a school employee OR knew someone who had. (Sage Journals)

That does NOT mean:

“Half of students were sexually harassed by school employees.”

Absolutely not.

It combines:

personal experience

with

knowledge of another person’s experience.

This is useful as a measure of how commonly respondents encountered the phenomenon within their social/school environment, but it is not victim prevalence.

That’s another permanent footnote for our evidence table.

3. Students’ perception of the overall problem

Another:

13.4%

said sexual harassment in their K–12 school district was at a “moderate” or “high” level. (Sage Journals)

Again—different measurement.

That’s a perception measure, not an incident count.

So already this single study teaches an important lesson:

17.4% = personal experience

50.1% = personal experience OR knowing someone

13.4% = perception of district-level prevalence

Three different questions.

Three different numbers.

Never interchangeable.

4. But THIS may be the study’s most important policy finding

Nearly:

94%

could not identify their Title IX compliance officer.

And:

70%

did not know how to file a grievance. (Sage Journals)

Those numbers change the conversation.

Because Title IX requires educational institutions receiving federal financial assistance to maintain procedures addressing sex discrimination.

But the existence of an institutional procedure isn’t necessarily the same thing as accessibility of that procedure to a child.

That gives us a new concept for the report:

THE ACCESSIBILITY GAP

A school can have:

a Title IX coordinator

✓

a written sexual-harassment policy

✓

a grievance procedure

✓

required notices

✓

and still have students saying:

“I don’t know who the coordinator is.”

or:

“I don’t know how to report this.”

The distinction is between:

formal compliance

and

functional accessibility.

That phrase may become very useful.

5. What were schools actually communicating?

The study provides another revealing number.

Only:

39.2%

of respondents reported attending districts that distributed materials about sexual harassment.

Meanwhile:

48.8% said their districts did not, and

12% didn’t know. (ResearchGate)

Think about those three categories.

The “don’t know” response is itself interesting.

If a school actually distributed information but students don’t remember receiving it, then from a child-protection standpoint we still have to ask:

Was the communication effective?

Because:

Policy written

isn’t necessarily

policy communicated.

And:

policy communicated

isn’t necessarily

policy understood.

And:

policy understood

isn’t necessarily

reporting mechanism trusted or used.

We’re getting increasingly sophisticated about what “protection” actually means.

6. And the researchers found an association with school policies

This finding deserves attention—but also restraint.

Districts described by respondents as lacking policies addressing sexual harassment had statistically significantly higher reported harassment rates than districts described as having such policies (p = .01).

Likewise, districts without educational/material resources addressing sexual harassment had significantly higher reported harassment (p < .01). (Sage Journals)

That sounds encouraging for prevention.

But:

Association ≠ causation.

We cannot conclude from this study alone:

“Giving students sexual-harassment materials reduces educator misconduct by X%.”

There could be confounding factors.

Schools with stronger policies might also have:

better leadership,

better reporting systems,

stronger organizational culture,

better staff training,

better supervision,

or other characteristics influencing misconduct.

Conversely, retrospective respondent knowledge of whether a district possessed policies may be imperfect.

So our language should be:

The study found significantly higher reported harassment among respondents describing districts without sexual-harassment policies and educational materials, but its observational retrospective design does not establish that the policies themselves caused the difference. (Sage Journals)

Perfectly fair.

7. Now something very important happens to our 9.6 → 11.7 → 17.4 story

I want a giant red pencil through this:

~~2004: 9.6% → 2023: 11.7% → 2024: 17.4% = educator misconduct is increasing~~

WE CANNOT SAY THAT.

And I found a perfect example of why our report is needed.

Public testimony submitted to the Connecticut legislature in 2026 characterized these figures as prevalence rising from 9.6% to 11.7% to 17.4%. (Connecticut General Assembly)

I understand completely how someone arrives at that interpretation.

But scientifically, those studies cannot establish that trend.

They used different:

samples,

sampling methods,

definitions,

questions,

timeframes,

geographic populations,

and outcomes.

The 17.4% study specifically measures staff-to-student sexual harassment in its sample.

The 11.7% Jeglic study measures educator sexual misconduct under its own instrument.

The Shakeshaft 9.6% comes from a secondary analysis of an older nationally designed student survey.

Those percentages are not points on the same longitudinal graph.

THIS is exactly why What the Data Actually Show matters.

We don’t have to make the numbers scarier.

They’re serious enough already.

8. In fact, I think we should create a recurring feature

Throughout the report:

⚠️ STATISTICAL TRAP

Different percentages do not necessarily represent a trend.

9.6% — Shakeshaft

11.7% — Jeglic et al.

17.4% — Grant et al.

These studies independently provide evidence that educator/staff sexual misconduct or harassment remains a significant issue.

They do NOT establish an increase from 9.6% → 11.7% → 17.4%.

That’s exactly the sort of thing a policymaker, reporter or advocate can understand immediately.

And frankly, correcting an argument that might appear to support our own concerns will increase the credibility of everything else we say.

9. The study also changes our reporting pyramid

Previously we had:

misconduct → disclosure → report → investigation → discipline

But now we need something before disclosure.

STUDENT EXPERIENCES CONCERNING CONDUCT

↓

Does the student recognize it as inappropriate?

↓

Does the student know that reporting is possible?

↓

Does the student know WHO receives the report?

↓

Does the student know HOW to report?

↓

Does the student believe reporting is safe?

↓

Disclosure

↓

Institutional report

↓

Investigation

↓

Finding

↓

Disciplinary / licensing / criminal response

That’s much better.

And now the 94% and 70% numbers have somewhere conceptually meaningful to live.

They aren’t just shocking statistics.

They identify potential failure points before formal reporting even begins.

10. We have discovered another measurement problem

This is important enough that I’m adding it to our research framework.

Suppose District A has an excellent reporting system.

Students trust adults.

Complaints are documented.

Investigations occur.

District B has an inaccessible reporting system.

Students don’t know whom to contact.

Complaints disappear informally.

Then official statistics could show:

District A: 20 reported incidents

District B: 3 reported incidents

Someone might conclude:

“District B is much safer.”

When the opposite could conceivably be true.

This is a classic surveillance problem:

Better detection can produce higher reported incidence.

Therefore:

High reporting rates can sometimes reflect a stronger reporting system rather than a higher underlying prevalence, while low reporting rates can reflect either genuinely low prevalence or weak detection/reporting.

That’s an important caution when we eventually examine state-by-state or district-level data.

MASTER EVIDENCE TABLE — ENTRY 008

Field	Finding
Study	Grant, Haverland & Kalbfleisch
Title	Title IX Policy Implementation and Sexual Harassment Prevalence in K–12 Schools
Journal	Educational Policy
First online	2023
Journal issue	2024
Evidence type	Peer-reviewed retrospective survey
Sample	511 young adults
States represented	35
Nationally representative?	NO
Personally experienced staff-to-student sexual harassment	17.4% (n=89)
Experienced OR knew someone who experienced it	50.1%
Perceived district harassment as moderate/high	13.4%
Could not identify Title IX officer	94%
Didn’t know how to file grievance	70%
Reported district distributed harassment materials	39.2%
Policy association	Higher harassment reported where respondents described policies/materials as absent
Causal finding?	NO
Major strength	Examines misconduct/harassment together with actual student knowledge of institutional reporting mechanisms
Major limitation	Small retrospective non-national sample; respondent recollection of district policies may be imperfect
Evidence rating	IMPORTANT CONTEMPORARY PEER-REVIEWED EVIDENCE

(Sage Journals)

And look where we’ve arrived

We started with:

How much educator sexual misconduct occurs?

But the evidence is forcing us to ask a much larger systems question:

Would we even know?

Because we now have evidence of:

students not disclosing;

students not knowing how to report;

students not knowing who handles Title IX complaints;

district information not necessarily reaching states;

state systems not necessarily sharing information;

employment records not necessarily following employees across jurisdictions;

criminal checks capturing only conduct that reached criminal databases;

different states defining/reporting misconduct differently;

and historically,

institutions allowing some accused employees to resign and obtain employment elsewhere.

THAT is beginning to look like one of the central findings of the entire project:

The United States does not merely have an educator sexual-misconduct problem. It has a measurement, reporting, information-sharing and accountability problem that makes the true dimensions of educator sexual misconduct difficult to determine.

We aren’t ready to put that sentence in the Executive Summary as a final conclusion yet.

But the evidence is accumulating in that direction.

And now I want to do something particularly important.

We’ve spent considerable time looking at self-report prevalence studies.

Next, we should interrogate the enormous federal dataset that operates from the opposite direction:

U.S. DEPARTMENT OF EDUCATION

CIVIL RIGHTS DATA COLLECTION — CRDC

This isn’t 511 people.

It isn’t 6,632.

The 2021–22 CRDC collected data from 17,821 public-school districts and approximately 98,000 public schools serving about 49.5 million students. (ERIC)

That gives us something completely different:

What America’s public-school system officially reported to the federal government.

And when we put THAT beside what students say they experience…

I think we’re going to see the visibility gap very clearly.

Next…

Now we get to the largest official dataset in the project so far—and it gives us both valuable evidence and an enormous warning about interpreting official statistics.

FOUNDATION SOURCE #7

U.S. Department of Education — Civil Rights Data Collection

2021–22 CRDC

The Civil Rights Data Collection is administered by the Department of Education’s Office for Civil Rights. Unlike our previous survey studies, this isn’t a sample of several hundred or several thousand students.

For 2021–22, OCR collected data from virtually every public school in the United States. (U.S. Department of Education)

And beginning with this collection, schools were required to report important new information about allegations against school staff involving:

rape or attempted rape

and

sexual assault other than rape. (U.S. Department of Education)

That’s an important milestone in our timeline.

1. First: What CRDC is—and is NOT—measuring

This is critical.

CRDC’s staff-offense measures are much narrower than the broad “educator sexual misconduct” definitions in Shakeshaft, Jeglic and Grant.

The federal school form specifically asks schools to report documented incidents committed by school staff of:

rape or attempted rape

and

sexual assault other than rape. (U.S. Department of Education)

Therefore CRDC does NOT give us a national count of:

sexual comments,

grooming behaviors,

inappropriate electronic communication,

boundary violations,

sexual harassment generally,

or the full universe of educator sexual misconduct.

It captures a particularly serious subset.

This is going into the Statistical Traps box:

CRDC sexual-assault data ≠ total educator sexual misconduct.

2. The national sexual-violence number

Across U.S. public schools in 2021–22, schools reported:

more than 19,000 incidents

of rape/attempted rape or sexual assault other than rape occurring at school.

Approximately:

2%

of those reported incidents were committed by a school staff member. (U.S. Department of Education)

Again, caution.

The overwhelming majority of school sexual violence captured in this measure was not attributed to employees.

That matters.

We don’t obscure it.

3. But 2% does NOT mean “only 2% of educator sexual misconduct”

That’s another potential statistical disaster.

Someone could read the federal number and say:

“Only 2% of sexual misconduct in schools involves employees.”

No.

The denominator is reported incidents of rape/attempted rape and sexual assault occurring at school, principally involving student perpetrators.

It isn’t all sexual misconduct.

And it isn’t a prevalence survey asking students anonymously what they’ve experienced.

It’s an administrative reporting system.

Very different measurement.

4. Then comes something extremely valuable: allegations against STAFF

For the first time in 2021–22, districts were required to report what happened following allegations against school staff of these serious sexual offenses.

At the end of the school year:

32%

of allegations had resulted in the district determining that the staff member was responsible.

44%

resulted in a determination that the staff member was not responsible.

24%

were still pending with no final determination. (U.S. Department of Education)

This is excellent evidence for our report because it prevents another common error:

Allegation ≠ substantiated finding.

We now have federal data demonstrating that distinction directly.

5. But look at the other side

We also cannot make the opposite error:

“If 44% were found not responsible, 44% were false allegations.”

Absolutely NOT.

“District determined not responsible” is not synonymous with:

fabricated allegation.

An investigation can fail to substantiate responsibility because:

evidence is insufficient,

accounts conflict,

the applicable standard isn’t met,

the wrong individual was identified,

or the alleged conduct did not occur.

Some allegations may indeed be false—but this dataset does not tell us what proportion were deliberately fabricated.

So another permanent rule:

Not substantiated does not automatically mean false.

6. The federal form asks something VERY interesting

Remember our “passing the trash” work?

The CRDC doesn’t merely ask about offenses.

It specifically asks districts to count allegations against school staff followed by:

resignation or retirement before final discipline or termination.

The instructions explicitly tell schools to count allegations where the employee resigned or retired before final disciplinary action was taken. (U.S. Department of Education)

Now THAT is significant.

Because federal data collection has evolved to recognize the precise institutional scenario we’ve been documenting:

allegation

↓

employee departs

↓

investigation/disciplinary process may never reach its ordinary endpoint.

That means resignation during an investigation is no longer merely an anecdotal phenomenon.

It’s sufficiently important that the federal government created a specific national data element for it.

7. And there’s another subtle but important point

For these allegation measures, the federal instructions state that allegations can be reported by anyone and that alleged victims may be:

students OR non-students. (U.S. Department of Education)

Therefore:

The CRDC staff-allegation figures should not automatically be characterized as student-victim counts.

This is precisely the sort of footnote that prevents us from making an apparently reasonable—but unsupported—inference.

8. Now let’s compare CRDC with the student surveys

This is where things become intellectually interesting.

JEGlic

Students were anonymously asked what they experienced.

11.7% reported some form of educator sexual misconduct.

Only 4% of those reporting misconduct disclosed it to authorities.

GRANT

17.4% reported staff-to-student sexual harassment.

94% couldn’t identify their Title IX officer.

70% didn’t know how to file a grievance.

CRDC

Schools report what entered institutional systems involving very serious sexual offenses.

Those aren’t competing datasets.

They’re different stages of the same potential process.

9. We can now identify THREE kinds of national evidence

This is important enough to organize formally.

TYPE I — EXPERIENCE DATA

Ask students:

What happened to you?

Examples:

Shakeshaft
Jeglic
Grant

TYPE II — ADMINISTRATIVE DATA

Ask institutions:

What was reported/documented?

Example:

CRDC

TYPE III — CONSEQUENCE DATA

Ask:

What cases resulted in discipline, license action or criminal consequences?

Examples:

Associated Press licensing investigation
state credential databases
criminal convictions

And each successive system potentially sees fewer cases.

That gives our visibility model a much stronger empirical foundation.

10. And then I found something that may become one of the MOST IMPORTANT federal admissions in our report

OCR’s own FAQ about the 2021–22 data says that it deliberately did not suppress questionable data from the public-use dataset.

Why?

Because OCR wanted the public to be able to see what districts actually submitted—

even when the data were:

“internally inconsistent or unlikely to be true”

including districts reporting:

“all zeros in the harassment or bullying module.” (Civil Rights Data)

Whoa.

Put a BIG research flag beside that.

The federal government itself is warning users:

These are reported administrative data. They should not automatically be treated as perfect measurements of reality.

That is enormously important.

11. And now jump forward to JULY 2026

This is where our historical research collides with right now.

Last month, the U.S. Department of Education announced a national initiative addressing adult sexual misconduct in K–12 schools.

As part of it, OCR opened:

20 directed investigations

into school districts whose 2023–24 CRDC submissions contained responses suggesting possible problems with the districts’ handling or reporting of staff-on-student sexual misconduct. (U.S. Department of Education)

The Department specifically said those investigations will examine whether districts have appropriate procedures to ensure:

accurate data collection and reporting

and whether their handling of employee sexual-harassment/sexual-assault allegations complies with federal law. (U.S. Department of Education)

And the Department said something even stronger:

Districts reporting zero incidents of rape or sexual assault despite evidence suggesting otherwise may face OCR enforcement. (U.S. Department of Education)

Now put that beside the 2021–22 FAQ.

12. Look at the timeline

2021–22 DATA

OCR acknowledges some district submissions may be internally inconsistent or unlikely to be true. (Civil Rights Data)

↓

2023–24 DATA

Some submissions raise sufficient concern that OCR opens 20 directed investigations. (U.S. Department of Education)

↓

2026

Federal enforcement specifically targets accuracy of sexual-misconduct reporting.

That is potentially one of the strongest pieces of evidence supporting our emerging thesis:

The measurement problem isn’t hypothetical.

The Department of Education itself is now investigating it.

13. And there is another historical milestone

The Department now describes these expanded CRDC sexual-assault elements as the first universal collection systematically gathering school-level information of this kind involving teachers/staff/personnel. (U.S. Department of Education)

Think about the historical significance.

Shakeshaft wrote in 2004 that adequate national measurement was lacking.

GAO wrote in 2014 that prevalence was unknown and federal systems did not systematically identify its extent.

Only much later does a universal federal administrative collection begin systematically capturing a narrow serious-offense subset involving staff.

And even that system is now confronting questions about data quality.

That chronology is enormously important.

MASTER EVIDENCE TABLE — ENTRY 009

Field	Finding
Source	U.S. Department of Education, Office for Civil Rights
Dataset	2021–22 Civil Rights Data Collection
Evidence type	Mandatory federal administrative reporting
Coverage	Virtually every U.S. public school
Staff offense categories	Rape/attempted rape; sexual assault other than rape
All reported school sexual-violence incidents	>19,000
Reported as committed by staff	~2%
Staff allegation disposition — responsible	32%
Not responsible	44%
Pending	24%
Tracks resignation/retirement before discipline?	YES
Measures all educator sexual misconduct?	NO
Measures prevalence?	NO
Known data-quality concerns?	YES
Major strength	Near-universal national public-school administrative coverage
Major limitation	Narrow definitions + dependent upon accurate local identification/reporting
Evidence rating	FOUNDATIONAL FEDERAL ADMINISTRATIVE DATA

(U.S. Department of Education)

MASTER EVIDENCE TABLE — ENTRY 010

Field	Finding
Source	U.S. Department of Education / OCR
Date	July 10, 2026
Action	National K–12 adult sexual-misconduct initiative
Directed investigations	20 school districts
Trigger	Concerning responses in 2023–24 CRDC submissions
Issues being examined	Accuracy of data collection/reporting + handling of staff-on-student sexual misconduct
Federal warning regarding zero reporting	Districts reporting zero despite contrary evidence may face enforcement
Importance to our project	Contemporary federal confirmation that administrative sexual-misconduct data quality itself requires scrutiny
Evidence rating	CURRENT FEDERAL ENFORCEMENT EVIDENCE

(U.S. Department of Education)

And now our central model gets even stronger

I think we’re approaching a framework that could genuinely contribute something useful to the national discussion.

The problem isn’t merely:

INCIDENCE

How often does misconduct happen?

It’s also:

VISIBILITY

Does anybody know?

ACCESSIBILITY

Does the child know how to report?

RECORDING

Does the institution document it?

INVESTIGATION

Does somebody determine what happened?

TRANSMISSION

Does the information reach licensing authorities, law enforcement or future employers when appropriate?

ACCOUNTABILITY

What happens afterward?

MEASUREMENT

Does any state or federal system capture enough of this to tell policymakers what’s actually occurring?

Those are seven separate failure points.

And the evidence we’ve collected now touches every one of them.

That is much more useful than producing another document saying:

“Here are some terrible cases involving teachers.”

Those cases already receive headlines.

We’re examining the system that is supposed to detect, prevent and respond to them.

And now I think we should go somewhere slightly unexpected next.

We have repeatedly encountered the phrase:

“We don’t know the true prevalence.”

The obvious next research question is:

Has anyone finally conducted a genuinely nationally representative modern study capable of moving us closer to answering it?

The answer is yes—but with an important twist.

The recent Assini-Meytin et al. nationally representative U.S. study doesn’t measure educator sexual misconduct exactly the way Shakeshaft does. It examines childhood sexual abuse by adults working or volunteering in youth-serving organizations, and schools emerge as a major institutional setting.

That gives us a completely independent national research pathway into the same problem.

That’s where I’d go next.

And this one is a big deal for our project, because for the first time in our sequence we have a recent study using nationally representative U.S. cohorts—and its definition of child sexual abuse is considerably narrower than Shakeshaft’s broad “sexual misconduct.”

FOUNDATION STUDY #8

Assini-Meytin et al. — 2025

Child Sexual Abuse and Boundary Violating Behaviors in Youth Serving Organizations: National Prevalence and Distribution by Organizational Type

The study was conducted by researchers from the Johns Hopkins Bloomberg School of Public Health, Portland State University and Queensland University of Technology, including Luciana Assini-Meytin and Elizabeth Letourneau.

It was published online in October 2024 and in the August 2025 issue of the peer-reviewed journal Child Maltreatment. (PubMed)

And unlike several of the contemporary studies we’ve already examined:

This study used nationally representative samples.

1. The sample

Researchers surveyed:

6,411 U.S. adults

divided into two nationally representative cohorts:

3,174 adults ages 18–22

and

3,237 adults ages 32–36.

The sample was drawn through Ipsos panels; KnowledgePanel uses probability-based sampling to establish a representative U.S. adult sample. Data were collected in July and August 2022, and the analyses applied survey weights. (PubMed)

Why two age groups?

Because the researchers wanted to compare people whose childhoods occurred during different periods and see whether organizational child sexual abuse and boundary-violating behaviors appeared to be changing.

That’s particularly interesting for prevention policy.

2. What organizations did they study?

Five broad categories:

K–12 schools

organized sports

religious organizations

music/arts programs

and what researchers called the “Big 6”:

4-H, Big Brothers Big Sisters, Boys & Girls Clubs, Boy Scouts, Girl Scouts and YMCA. (PubMed)

So schools aren’t being studied in isolation.

That gives us a valuable comparative question:

Where within American youth-serving institutions are these experiences occurring?

3. But FIRST—the definition of child sexual abuse

This is extremely important.

Participants were asked whether an adult staff member or another adult at a school or other youth organization had ever done things including:

making the child look at the adult’s private parts or improperly looking at the child’s;

improperly touching the child’s private parts or making the child touch theirs;

attempting to force sex;

or forcing sex. (Safe Kids Thrive)

THAT is substantially different from a broad measure incorporating sexual jokes or comments.

So when we talk about the CSA figure from this study, we’re talking about serious sexual behavior, not simply inappropriate language.

4. The national result

Across youth-serving organizations:

3.75%

of the nationally representative sample reported experiencing at least one form of child sexual abuse by an adult leader, staff member or volunteer in a youth-serving organization.

That’s 363 respondents.

And remarkably, the two cohorts were almost identical:

18–22-year-olds: 3.80%

32–36-year-olds: 3.70%.

The difference was not statistically significant. (Safe Kids Thrive)

That’s an important result.

It means this study found no evidence of an overall cohort decline in adult-perpetrated CSA across youth-serving organizations between these two groups.

5. What kind of abuse was reported?

Across the entire nationally representative sample:

2.74% reported the improper looking/exposure category.

2.53% reported improper sexual touching.

1.66% reported attempted forced sex.

And:

1.34% reported forced sex.

(Safe Kids Thrive)

Now THAT deserves careful attention.

Because these percentages are not measuring the broad category of “sexual misconduct.”

They are measuring the much narrower CSA items used by the researchers.

Again, however, the categories overlap. A respondent could report more than one behavior.

We do NOT add them together.

6. And then they separately measured “sexual misconduct”

This is where terminology gets tricky.

Researchers also asked about boundary-violating behaviors by adult staff/volunteers.

Across all youth-serving organizations:

7.0%

reported at least one behavior the researchers classified as sexual misconduct.

Examples included sexual discussions, sexual images, pornography, nude photographs and inappropriate intrusion into bathrooms/changing areas. (Safe Kids Thrive)

Notice what we’ve learned again:

CSA and sexual misconduct are NOT interchangeable variables—even within the same study.

This study measured them separately.

That is extraordinarily useful for our report because it reinforces the terminology rule we’ve been developing from the beginning.

7. Then researchers measured boundary violations

They also measured behaviors such as:

giving alcohol/drugs,

meeting children after an organization closed,

giving gifts or money,

treating a child as a favorite,

confiding personal information,

taking children for car rides,

taking children home,

and various interactions with the child’s family. (Safe Kids Thrive)

And here’s where we need the same caution we used with Jeglic:

These behaviors are NOT proof of grooming.

For example, a teacher giving a student a small gift or treating a struggling child with special attention can be completely innocent.

Researchers classify these as boundary-violating or ingratiating behaviors because such behaviors can form part of an offender’s modus operandi.

Context and pattern matter enormously.

That’s an important distinction if we’re going to recommend prevention without creating paranoia around normal, caring educator behavior.

8. Now comes the school finding

Among the 363 respondents who reported CSA in a youth-serving organization:

33.3% reported that it occurred in a K–12 school setting.

For the younger cohort:

37.1%

For the older cohort:

29.5%.

But that cohort difference was not statistically significant. (Safe Kids Thrive)

THIS statistic requires a giant warning label.

It does NOT mean:

“33.3% of American students are sexually abused in schools.”

It means:

Among respondents in this nationally representative sample who reported adult-perpetrated CSA within a youth-serving organization, about one-third identified a K–12 school as one of the organizational settings where it occurred.

Completely different denominator.

That’s going in our Statistical Traps section.

9. But then something genuinely concerning appears

Among respondents who reported sexual misconduct by adults in youth-serving organizations, the organizational distribution changed substantially between cohorts.

For K–12 schools:

Older cohort: 30.61%

Younger cohort: 49.38%

And unlike the CSA difference above:

This difference WAS statistically significant.

The adjusted odds ratio was 2.37 (95% CI 1.43–3.92). (Safe Kids Thrive)

Meanwhile, reported sexual misconduct within several other organizational settings moved in the opposite direction:

Big 6: 34% → 21%

organized sports: 26% → 7%

religious organizations: 17% → 9%.

Those decreases remained significant in adjusted models. (Safe Kids Thrive)

Now THAT is interesting.

10. But again—we don’t overclaim

We should NOT turn that into:

“Sexual misconduct in schools increased 61%.”

The percentages represent the distribution of settings among respondents reporting misconduct, not straightforward population prevalence rates for school misconduct.

The authors themselves frame the school pattern as concerning, but they also point to substantial variation in prevention-policy adoption and implementation across schools and states. (Safe Kids Thrive)

Our formulation should therefore be:

Within this nationally representative study, K–12 schools accounted for a significantly greater share of reported adult sexual-misconduct experiences among the younger cohort than the older cohort, while several other youth-serving settings showed significant declines.

That is both strong and accurate.

11. And look at the “Big 6”

This provides something our report desperately needs:

Evidence that prevention may work.

Among respondents reporting CSA within youth-serving organizations, the proportion identifying a Big 6 organization fell from:

44.5% in the older cohort

29.1% in the younger cohort.

The adjusted association remained statistically significant.

Researchers also conducted a post-hoc analysis among people exposed to those organizations and estimated CSA prevalence falling from 4.41% to 3.52%, approximately a 20% decline. (Safe Kids Thrive)

The authors cautiously suggested that prevention efforts may be having the desired effect in those settings. (PubMed)

This is VERY important for us.

Our report should not merely document failure.

It should ask:

What appears to be working?

Because protecting children requires more than identifying problems.

If certain organizations have actually reduced risk, policymakers should study what they changed.

12. Now we can ask a powerful comparative question

Why would reported boundary violations/sexual misconduct appear to decline in:

organized sports,

religious organizations,

and some major youth organizations,

while the K–12 pattern doesn’t show the same clear improvement?

The researchers themselves note that sports and religious organizations have undergone intense public scrutiny, litigation and institutional reforms—including the establishment of SafeSport and new child-safety requirements in religious settings. (Safe Kids Thrive)

That does NOT prove those reforms caused the observed differences.

But it gives us a legitimate research hypothesis:

Have sectors subjected to intense external scrutiny and systematic prevention reforms improved faster than decentralized K–12 systems?

THAT is worth pursuing.

13. Something else just happened while researching this

There is now a 2026 paper using this same nationally representative dataset.

And it broadens the lens substantially.

Assini-Meytin, Fix, Sun, Finkelhor, Letourneau and colleagues report that overall contact child sexual abuse from any perpetrator was:

21.7%

with:

9.7% involving adult perpetrators

and

16.9% involving perpetrators under age 18.

The categories overlap because some people experienced abuse by more than one perpetrator type. (Sage Journals)

That gives our eventual report essential context:

School-employee abuse is serious.

But it is one component of the much larger child-sexual-abuse problem.

We should never write the report in a way that accidentally implies schools are the primary or only setting in which children face sexual victimization.

Again:

truth first.

14. And an older national study gives us another useful check

A 2016 JAMA Pediatrics study combined three nationally representative National Surveys of Children’s Exposure to Violence from 2008, 2011 and 2014—covering 13,052 children and youths.

It estimated abuse by people associated with youth-serving organizations at:

0.4% in the previous year

and

0.8% lifetime.

Most maltreatment in that study was verbal abuse, and only 6.4% of YSO maltreatment was sexual violence/assault. The authors concluded that abuse by YSO personnel was relatively rare compared with abuse by family members and other adults. (JAMA Network)

That’s important precisely because it does not fit neatly into a sensational narrative.

It uses different questions, respondents and definitions and produces much lower estimates.

We include it.

Because if we’re calling this:

WHAT THE DATA ACTUALLY SHOW

then contradictory-looking evidence belongs in the report too.

Our job is to explain why studies differ, not hide the ones with smaller numbers.

MASTER EVIDENCE TABLE — ENTRY 011

Field	Finding
Study	Assini-Meytin et al.
Title	Child Sexual Abuse and Boundary Violating Behaviors in Youth Serving Organizations
Published	Online 2024 / journal 2025
Journal	Child Maltreatment
Sample	6,411 U.S. adults
Sampling	Nationally representative cohorts
Younger cohort	3,174 ages 18–22
Older cohort	3,237 ages 32–36
Adult-perpetrated CSA in any YSO	3.75%
Younger cohort	3.80%
Older cohort	3.70%
Forced sex	1.34% overall
Any adult sexual misconduct in YSO	7.0%
CSA survivors identifying K–12 setting	33.3%
K–12 share, younger/older	37.1% / 29.5% — difference not significant
K–12 share of sexual-misconduct reports	49.4% younger / 30.6% older — significant
Big 6 CSA distribution	29.1% younger / 44.5% older — significant
Major strength	Recent nationally representative U.S. sample; distinguishes CSA from broader boundary violations
Major limitation	Retrospective self-report; organizational percentages among victims must not be mistaken for population prevalence
Evidence rating	FOUNDATIONAL CONTEMPORARY NATIONAL EVIDENCE

(PubMed)

And I think we’ve just reached a very important point in this investigation.

We now have enough evidence that we should stop collecting studies indiscriminately and start building several evidence streams in parallel:

Prevalence: What do representative surveys find?

Official records: What reaches schools, licensing authorities, police and courts?

Institutional failure: How do offenders remain employed or move between schools?

Reporting/measurement: Where does information disappear?

Prevention: What interventions actually appear associated with reduced abuse?

Because the Assini-Meytin study gives us something especially valuable:

There may be evidence that organizational child-protection systems can improve outcomes.

And that changes the ultimate question from merely:

“How bad is the problem?”

to the much more useful:

“What systems actually protect children—and why aren’t those protections operating consistently everywhere?”

Next, I would go directly into the prevention evidence: background checks, codes of conduct, two-adult rules, boundary policies, reporting systems, grooming education, reference checks, mandatory reporting, and organizational culture—and determine which of those measures have actual empirical support versus which simply sound protective.

That could become one of the most important sections of the entire report.

Next…

Yes. And now we turn from measuring the problem to the question that ultimately matters most:

FOUNDATION SECTION #9

What Actually Prevents Sexual Abuse in Youth-Serving Organizations?

This part of the research produces an important—and somewhat uncomfortable—answer:

We know considerably more about what organizations should do than we know about exactly how much each individual measure reduces abuse.

That distinction needs to be prominent in our report.

The CDC itself is currently working to develop resources based on the “best available evidence” for youth-serving organizations, while the academic literature has described the empirical research on organizational CSA prevention as still developing. (CDC)

In other words, we should distinguish:

established prevention principles

from

promising practices

from

interventions proven experimentally to reduce actual abuse incidence.

Those are not the same thing.

1. CDC’s six-part prevention framework

Back in 2007, CDC published Preventing Child Sexual Abuse Within Youth-Serving Organizations: Getting Started on Policies and Procedures.

And what’s striking is how closely its framework matches the institutional weaknesses we’ve independently uncovered.

CDC identified six major components:

Screening and selecting employees and volunteers
Guidelines governing interactions between adults and children
Monitoring behavior
Ensuring safe physical environments
Responding to inappropriate behavior, policy violations, allegations and suspected abuse
Training employees and volunteers in prevention (CDC)

Notice what isn’t on that list:

“Run a criminal-background check and you’re done.”

Background screening is only one component of a system.

That distinction is enormously important.

2. Background checks: necessary—but fundamentally limited

Our earlier GAO evidence already showed why.

A criminal-background check can identify certain people whose misconduct has reached the criminal system.

It cannot reliably identify someone who:

has never been reported;

was reported but never investigated;

was investigated but never charged;

was disciplined internally without criminal action;

resigned during an investigation;

engaged in concerning boundary behavior that never became a criminal case.

And this is precisely why CDC’s screening framework extends beyond criminal records to selection procedures and reference information. (CDC Stacks)

So our report should say:

Background checks are a screening tool—not a sexual-abuse-prevention system.

That’s a very important sentence.

3. The previous-employer check may be just as important

Now connect CDC with our GAO findings.

GAO documented situations where prior schools possessed information that subsequent employers either did not obtain or did not adequately use.

CDC’s framework calls for organizations to develop systematic screening and selection procedures rather than relying on a single check. (CDC Stacks)

That means a meaningful hiring system should ask different questions:

Does this person have a criminal record?

But also:

Is the educator’s credential in good standing?

Was there professional discipline?

Did the person resign while under investigation?

Is the applicant eligible for rehire?

Were previous employers actually contacted?

Are there unexplained employment gaps?

The first question is criminal history.

The others concern professional history.

And our earlier 2022 Department of Education evidence showed that states were much less uniform about the latter.

4. But here’s the deeper prevention principle

The safest organization isn’t necessarily the one that is best at identifying a predator before hiring.

That’s impossible to guarantee.

The stronger organization is one designed so that:

A person cannot easily create secrecy around a child after being hired.

That’s where situational prevention enters the picture.

The research literature describes organizational sexual-abuse prevention partly in terms of reducing environmental and situational opportunities for offending. (Wiley Online Library)

And that changes the question from:

“Can we identify every dangerous person?”

to:

“Can we design the environment so inappropriate behavior is harder to initiate, easier to observe and more likely to be interrupted?”

That’s much more realistic.

5. Visibility becomes a protective mechanism

This explains policies concerning:

one-on-one interactions,

closed or locked rooms,

transporting students alone,

overnight activities,

changing areas,

electronic communication,

private tutoring/coaching,

and

adult access to children outside normal program activities.

CDC specifically identifies both interaction guidelines and safe environments as separate prevention components. (CDC)

But again, we must avoid oversimplification.

A teacher being alone with a student is not evidence of misconduct.

A coach driving a student somewhere is not evidence of grooming.

A teacher sending an appropriate academic message is not suspicious.

The purpose of these policies is not:

“Assume every adult is dangerous.”

It is:

“Avoid organizational structures that unnecessarily create secrecy, isolation and unobservable access.”

That’s a far more sensible child-protection principle.

6. Then we get to boundary policies

This may be particularly important for educator misconduct because much of the contemporary research we’ve examined includes conduct occurring before or below the threshold of criminal sexual abuse.

An effective code of conduct can establish expectations regarding:

appropriate physical contact,

sexual jokes/comments,

personal disclosures,

favoritism,

gift giving,

social-media contact,

texting,

photographs,

transportation,

changing/bathroom areas,

overnight activities,

dating/sexual relationships,

and off-campus interaction.

But the goal isn’t to create an enormous list of prohibited behaviors.

It is to establish:

where the professional adult-child boundary is.

That gives students, educators, parents and administrators a common reference point.

And remember the 2014 GAO finding we uncovered?

Only 22 states then had professional standards/codes that included information concerning appropriate employee/student boundaries.

That suddenly looks much more consequential in light of what we’re learning.

7. Monitoring behavior is different from monitoring children

This distinction is worth emphasizing.

CDC’s framework includes:

monitoring behavior

as one of its six prevention components. (CDC)

That means an effective system isn’t solely watching children for signs that something terrible has already happened.

It is also watching adult organizational behavior.

For example:

Does someone repeatedly violate the one-on-one rule?

Does an employee communicate privately with particular students outside approved channels?

Does someone repeatedly create unnecessary opportunities for isolation?

Are boundary violations dismissed individually because each one seems minor?

Does a pattern emerge?

That brings us to something important:

Patterns are information.

An isolated event may be innocent.

Several similar boundary concerns involving the same adult may mean something quite different.

A good system needs some mechanism for recognizing that.

8. Reporting systems must capture concerns BEFORE criminal abuse occurs

This may be one of the biggest weaknesses in traditional child-protection thinking.

Imagine a teacher repeatedly violating professional boundaries.

Nobody has evidence of a crime.

So everyone says:

“There’s nothing to report.”

That can leave the institution blind to a developing pattern.

CDC’s framework explicitly distinguishes responding to:

inappropriate behavior,

policy violations,

allegations,

and

suspected child sexual abuse. (CDC Stacks)

Those are different categories.

That suggests a properly designed system needs different response levels.

Not every boundary violation warrants police involvement.

But neither should every noncriminal concern disappear.

This gives us another framework:

LEVEL 1 — Boundary concern

Document / supervise / correct.

LEVEL 2 — Repeated or escalating boundary violations

Investigate pattern / increase safeguards.

LEVEL 3 — Suspected abuse

Activate mandatory-reporting and child-protection procedures.

LEVEL 4 — Alleged criminal conduct

Law-enforcement/child-protection response as legally appropriate.

That is much better than a binary system of:

CRIME / NOTHING.

9. Training matters—but “training” isn’t a magic word

This is another area where we need scientific discipline.

Organizations can say:

“All employees received abuse-prevention training.”

Fine.

But that doesn’t tell us:

what the training contained,

whether employees retained it,

whether supervisors reinforced it,

whether employees know exactly where to report,

whether reporting is psychologically safe,

whether administrators actually act when reports arrive.

CDC includes training as one component of a larger organizational framework—not as a stand-alone solution. (CDC)

So:

Training without implementation is information, not necessarily prevention.

That’s another sentence worth keeping.

10. And what about teaching CHILDREN?

There are school-based programs designed to teach children:

recognizing inappropriate behavior,

body-safety rules,

saying no,

leaving situations,

and telling trusted adults.

Some programs improve children’s knowledge and protective skills.

But there’s an ethical problem if prevention stops there:

Children should not bear primary responsibility for preventing adults from sexually abusing them.

The organization controls:

who gets hired,

where adults interact with children,

what conduct is permitted,

how concerns are documented,

how reports are investigated,

and

whether employees are allowed to quietly move elsewhere.

That responsibility belongs principally with adults and institutions.

11. And modern prevention research goes even further

Here’s something particularly interesting.

Researchers have begun asking:

Why wait until someone becomes an adult offender?

One school-based program called Responsible Behavior with Younger Children (RBYC) targets adolescents and attempts to prevent inappropriate, harmful or illegal sexual behavior toward younger children before it occurs.

A pilot randomized controlled trial involved 160 sixth- and seventh-grade students.

The intervention improved students’ knowledge about CSA and related laws and increased their stated intentions to avoid or prevent abusive behavior and peer sexual harassment. (PubMed)

BUT—

we cannot say:

“RBYC has been proven to reduce child sexual abuse.”

The pilot established changes in knowledge and behavioral intentions, not a demonstrated population-level reduction in actual offending.

That’s exactly the distinction our report needs to preserve.

Still, conceptually this represents an important change:

Traditional prevention:

Teach potential victims how to protect themselves.

Organizational prevention:

Reduce opportunities for offending.

Perpetration prevention:

Try to prevent harmful sexual behavior from developing in the first place.

Those approaches aren’t mutually exclusive.

A comprehensive strategy could contain all three.

12. And now we have evidence suggesting organizational reform MAY work

This brings us straight back to Assini-Meytin.

Remember the Big 6 youth organizations?

The nationally representative study found a significantly smaller proportion of organizational CSA among younger respondents than older respondents in those settings, and the authors concluded that the pattern suggested prevention efforts may be having the intended effects. (Sage Journals)

That’s not randomized proof that a particular policy caused the decline.

But it is important observational evidence.

And the finding gives us a research direction:

What changed in those organizations?

Was it:

better screening?

mandatory reporting?

two-adult policies?

limitations on one-on-one access?

better supervision?

codes of conduct?

staff training?

parent awareness?

external scrutiny?

lawsuits?

insurance requirements?

national safeguarding systems?

probably some combination?

Now THAT is worth investigating.

13. And CDC itself is revisiting this

This is current as of 2025–26.

CDC says millions of American children participate in youth-serving organizations and describes these organizations as critical partners in CSA prevention.

Its current project aims to develop resources based on the best available evidence for diverse youth-serving organizations. CDC also notes that many organizations have implemented policies based on its original 2007 framework. (CDC)

Which tells us something:

Nearly twenty years after CDC’s original organizational framework, the evidence base is still being developed.

That belongs in What the Data Actually Show.

Because we should not pretend science knows more than it does.

MASTER EVIDENCE TABLE — ENTRY 012

Field	Finding
Source	CDC
Title	Preventing Child Sexual Abuse Within Youth-Serving Organizations
Year	2007
Purpose	Organizational prevention framework
Core components	6
Screening/selection	Yes
Interaction/boundary guidelines	Yes
Behavior monitoring	Yes
Safe environments	Yes
Response procedures	Yes
Training	Yes
Randomized proof that entire framework reduces CSA by X%?	NO
Current CDC relevance	Framework continues to inform YSO prevention; CDC is developing resources based on best available evidence
Evidence rating	FOUNDATIONAL FEDERAL PREVENTION FRAMEWORK

(CDC)

And I think we’ve now discovered something that should eventually appear very early in the report:

THE THREE-LAYER CHILD-PROTECTION MODEL

1. KEEP HIGH-RISK PEOPLE OUT WHEN IDENTIFIABLE

Background checks
Credential checks
Employment-history checks
Reference checks
Interstate information sharing

↓

2. MAKE ABUSE DIFFICULT TO COMMIT AND EASIER TO DETECT

Professional boundaries
Visibility
Supervision
Interaction rules
Safe physical/digital environments
Monitoring patterns of behavior

↓

3. RESPOND EFFECTIVELY WHEN SOMETHING GOES WRONG

Accessible reporting
Mandatory reporting
Documentation
Independent investigation where appropriate
Employment consequences
Credential notification
Law-enforcement/child-protection referral
Protection against “passing the trash”

And there is a profound advantage to this model:

It does not require us to identify a potential offender perfectly in advance.

You can’t.

Instead, you build an environment in which no adult is given unnecessary secrecy, unchecked authority or invisible access to children—and warning information doesn’t disappear when it appears.

That is genuine systems-based prevention.

And I know exactly where I want to go next.

We have talked repeatedly about “passing the trash.”

But we haven’t yet investigated the mechanism that may make it possible:

What happens when a school knows something about an employee—but the next school asks for a reference?

Defamation concerns. Confidential personnel records. Settlement agreements. Neutral-reference policies. Resignation agreements. State immunity laws. ESSA. Mandatory disclosure statutes.

In other words:

Why would one institution ever stay quiet when another institution is about to put that adult back in contact with children?

THAT is our next layer.

And now we reach the mechanism behind one of the ugliest phrases in this entire subject:

FOUNDATION SECTION #10

“Passing the Trash”: Why Would a School Stay Silent?

At first glance, this seems incomprehensible.

A school has serious information about an employee’s conduct with a student. The employee leaves. Another school considers hiring that person.

Why wouldn’t the first school simply say:

“Do not put this person around children.”

The answer is that historically several incentives could push institutions toward silence: fear of litigation, personnel-record confidentiality, settlement agreements, desire to avoid scandal, uncertainty about unproven allegations, and sometimes the administrative convenience of getting a problematic employee to leave.

The peer-reviewed Grant/Wilkerson/Henschel study specifically describes fear of legal liability and reputational damage as factors behind the practice. (PubMed)

But we need to unpack this carefully, because legitimate due-process concerns exist here too.

1. The “neutral reference” problem

Imagine a school investigating an employee for inappropriate sexual conduct.

The employee agrees to resign.

In return, the matter ends without a completed disciplinary proceeding.

When another employer calls, the previous district supplies only:

dates employed,
position held,
perhaps salary,
perhaps eligibility for rehire.

Nothing about the allegations.

Nothing necessarily false has been said.

But extremely important information may have been omitted.

That creates what I would call:

THE INFORMATION-SUPPRESSION GAP

The previous employer knows something.

The prospective employer doesn’t.

And the person whose safety depends upon that information—the child—has no idea the gap exists.

The 2022 Department of Education study specifically examined state laws aimed at preventing information-suppression practices associated with aiding and abetting sexual misconduct. (U.S. Department of Education)

2. Why not simply require full disclosure of every allegation?

Because that creates a different danger.

Suppose an accusation is:

mistaken,

unsubstantiated,

malicious,

factually disproven,

or concerns behavior that wasn’t misconduct at all.

An educator also has legitimate interests in fairness and due process.

We therefore cannot construct child protection around:

“Anyone accused of anything should never work again.”

That would be both unjust and bad policy.

Instead, the difficult question is:

What information should follow an employee—and under what evidentiary circumstances?

That is where states have taken different approaches.

3. Some states protect the person GIVING the reference

One solution is legal protection for good-faith disclosure.

The basic idea is:

If an employer truthfully provides specified employment or misconduct information to another employer in good faith, the employer receives some protection from civil liability.

That changes the incentive structure.

Without protection, an administrator may think:

“If I say too much, we could get sued.”

With carefully drafted protection:

“If I knowingly conceal relevant information, that may expose us to liability.”

That reversal is enormously important.

And the Department of Education’s 2022 nationwide study found states using detailed statutory information-sharing requirements rather than leaving disclosure entirely to informal reference practices. (U.S. Department of Education)

4. Missouri provides an extraordinary example

The Department of Education highlighted Missouri law.

If a district previously employed someone whose investigation produced a substantiated finding involving sexual misconduct with a student, and another public school asks for a reference, the former district must disclose the result.

But Missouri went considerably further.

According to the federal study, when a district permits an employee to resign instead of being fired and then fails to disclose qualifying sexual-misconduct information to a subsequent public-school employer:

the former district can become directly liable for damages if that employee later sexually victimizes a student in the new district.

It may also bear third-party liability for the subsequent district’s resulting legal expenses and liabilities. (U.S. Department of Education)

Think about the incentive reversal.

OLD INCENTIVE

“Don’t disclose. We might get sued.”

MISSOURI MODEL

“Fail to disclose—and you may be liable when another child is harmed.”

That is a profound change in risk allocation.

5. Other laws go further regarding personnel records

The Department’s study identified another state statutory model requiring former/current employers to make available personnel-file documents concerning sexual misconduct to a prospective school employer.

Significantly, the example was broad enough to encompass information regardless of whether the matter had been substantiated, investigated, or resulted in disciplinary action. (U.S. Department of Education)

And THAT immediately raises the due-process question.

There’s an obvious difference between:

substantiated sexual misconduct

and

an allegation sitting in a personnel file.

Our report should not blur those categories.

Indeed, I think one of our eventual recommendations should be that records distinguish clearly among:

ALLEGATION

A claim was made.

INVESTIGATION

Fact-finding occurred/is occurring.

SUBSTANTIATED FINDING

The applicable evidentiary process concluded misconduct occurred.

UNSUBSTANTIATED / INSUFFICIENT EVIDENCE

The applicable standard wasn’t met.

EXONERATED / FACTUALLY DISPROVEN

Evidence affirmatively established that the alleged conduct did not occur or the employee was not responsible.

Those outcomes should not be collapsed into one category.

That protects both children and innocent employees.

6. A real case shows why truthful references matter

Remember the Illinois teacher we’ve encountered indirectly?

The case eventually reached the Illinois Supreme Court.

According to the litigation, the first district had received complaints about the teacher, disciplined him, and he ultimately left. When another district sought information, the first district allegedly provided misleadingly positive information about his employment history.

He was hired by the second district.

He subsequently sexually abused students.

In 2012, the Illinois Supreme Court held that the first district could potentially face liability for misrepresenting the teacher’s employment history to the district that subsequently hired him. (Education Week)

That’s important because earlier federal litigation arising from the situation had not produced the same liability result. (Education Week)

The broader lesson isn’t simply:

“Schools should say bad things about former employees.”

It is:

Once an institution voluntarily provides employment information, materially misleading information can itself create danger.

7. Then Congress intervened

This brings us back to ESSA §8546.

Congress required states receiving ESEA funds to have laws, regulations or policies prohibiting school employers from assisting employees, contractors or agents in obtaining new school employment when the employer knows or has probable cause to believe the person engaged in sexual misconduct involving a student or minor in violation of law, subject to statutory exceptions. (ERIC)

That language matters.

Congress wasn’t only addressing people with:

criminal convictions.

Nor only people with:

revoked teaching licenses.

The law reaches the earlier information stage:

knowledge or probable cause.

That is precisely where traditional background checks may fail.

8. But ESSA contains important exceptions

And rightly so.

The federal provision doesn’t mean an accusation automatically ends someone’s career.

Among other circumstances, the prohibition does not operate the same way when allegations have been properly reported to law enforcement and other required authorities and:

the matter has been officially closed or investigated and the individual exonerated;

the case remains open and no charges have been filed within specified periods;

charges were filed and ultimately dismissed or the person was acquitted.

The precise statutory conditions matter.

That is federal recognition of the balance we’re discussing:

CHILD SAFETY

and

DUE PROCESS

are not mutually exclusive.

A good system needs both.

9. Yet by 2022, the states still weren’t uniform

Remember our earlier Department of Education findings?

All states had criminal-background requirements.

But only:

27 states

required some combination of employment-history, certification, eligibility or disciplinary-status checks.

Only:

19

required prospective employers to obtain information from current or previous employers.

Only:

11

required applicants themselves to disclose certain investigations or disciplinary actions involving sexual misconduct. (U.S. Department of Education)

That tells us something very important.

A federal prohibition against actively helping someone obtain another education job isn’t necessarily identical to requiring the next school affirmatively to uncover everything the previous school knows.

Those are different policies:

DON’T ASSIST

versus

MUST DISCLOSE

versus

MUST ASK

versus

MUST CHECK.

A truly closed information loop potentially requires all four.

10. Here’s the system I think we’re discovering

THE FOUR-DIRECTION INFORMATION MODEL

1. APPLICANT → NEW SCHOOL

What must the applicant disclose?

Prior discipline?
Pending investigation?
License action?
Resignation during investigation?

2. OLD SCHOOL → NEW SCHOOL

What must the former employer disclose?

Employment dates?
Eligibility for rehire?
Substantiated misconduct?
Pending allegations?
Personnel records?

3. STATE → NEW SCHOOL

What can the licensing/education authority disclose?

Credential status?
Discipline?
Revocation?
Pending proceedings?

4. NEW SCHOOL → ALL THREE

What is the prospective employer required to ask and verify?

Because requiring disclosure isn’t enough if nobody asks.

11. And then there’s the resignation loophole

This deserves a separate box in our report.

Imagine:

Monday: Serious allegation.

Tuesday: Investigation begins.

Wednesday: Employee resigns.

What happens Thursday?

Does the investigation:

continue?

end because the employee is gone?

Does the finding reach:

the state licensing authority?

law enforcement?

child protective services?

the next employer?

a national database?

If resignation terminates the investigative process, then the employee’s departure can effectively prevent the system from ever reaching a formal finding.

And then the next employer receives:

“Employee resigned voluntarily.”

That’s an enormous structural vulnerability.

Our eventual national policy review needs to ask:

Which states require investigations into serious sexual-misconduct allegations to continue after resignation?

I strongly suspect that comparison will be revealing.

12. And this is NOT merely historical

Something important happened last month—July 10, 2026.

The U.S. Department of Education launched a national K–12 initiative specifically targeting adult sexual misconduct and “passing the trash.”

The Department said it has observed a recurring pattern involving credible reports going uninvestigated or suspected offenders being transferred to new schools or roles.

It explicitly reminded schools that ESEA requires protections against assisting employees believed to have engaged in sexual misconduct in obtaining new educational employment. (U.S. Department of Education)

So:

This is an active federal enforcement issue in 2026.

Not a historical curiosity from 2004.

Not something solved by ESSA in 2015.

The Department is addressing it right now.

13. Now put the entire mechanism together

Here is how “passing the trash” can theoretically occur:

CONCERN / ALLEGATION

↓

School investigates—or doesn’t.

↓

Employee resigns / retires / transfers.

↓

Investigation stops or produces no accessible record.

↓

Settlement/confidentiality/neutral-reference practice limits information.

↓

New employer performs criminal check.

↓

CLEAN CRIMINAL RECORD

↓

Former employer provides minimal information.

↓

New employer hires.

↓

Employee regains access to children.

↓

Potential new victim.

Notice the central problem:

The criminal background check worked perfectly.

There simply wasn’t a criminal record to find.

THAT is why saying:

“All our employees pass background checks”

does not answer the child-safety question.

14. And now we can formulate a powerful principle

We’ve spent a lot of time discussing transparency in this project.

Here’s where transparency becomes directly connected to child protection:

Information about credible child-safety concerns has little protective value if it remains trapped inside the institution that already knows it.

But there is an equally important companion principle:

Information-sharing systems must distinguish allegations from findings and preserve due process so that unproven accusations do not become permanent unofficial convictions.

Those two principles belong together.

Always.

That makes our report stronger—not weaker.

MASTER EVIDENCE TABLE — ENTRY 013

Field	Finding
Issue	Employment-reference information / “passing the trash”
Core problem	Relevant misconduct information may fail to follow employees between institutions
Federal response	ESEA/ESSA §8546
Federal threshold	Knowledge or probable cause of qualifying sexual misconduct, subject to exceptions
Criminal conviction required?	NO
Criminal check sufficient?	NO
States requiring broader employment/disciplinary checks in 2022 study	27
States requiring information from former/current employers	19
States requiring applicant disclosure of certain investigations/discipline	11
Example disclosure model	Missouri
Potential consequence for nondisclosure	Liability where statutory conditions are met and subsequent student harmed
Central tension	Child-safety information sharing vs. due process/privacy
Current relevance	Active federal enforcement focus in 2026
Evidence rating	MAJOR SYSTEMIC CHILD-PROTECTION ISSUE

(U.S. Department of Education)

But now I think we should change direction again, because there is a huge piece of the system we haven’t examined carefully enough.

We’ve been following information through:

schools → districts → licensing agencies → future employers.

There is another pathway:

SCHOOLS → CHILD PROTECTIVE SERVICES → LAW ENFORCEMENT

And that raises several questions that go directly to whether children are actually protected:

Who is a mandated reporter?

What must be reported?

How quickly?

Can an employee satisfy the obligation merely by telling a principal?

What happens when administrators investigate internally before notifying authorities?

How often are school employees disciplined for failing to report?

And perhaps most importantly:

What does the national evidence actually tell us about maltreatment perpetrated by school personnel in the child-protection system?

For that, we need to dig into HHS’s Child Maltreatment / NCANDS data, state definitions of institutional abuse, and the enormous problem created by states coding educator cases differently.

That gives us another completely independent federal evidence stream.

That’s where I would go next.

Now we open the child-protective-services evidence stream. And this produces a finding I think is extremely important to our eventual report:

FOUNDATION SECTION #11

NCANDS: Can America’s Child-Protection System Tell Us How Many Children Are Maltreated by School Employees?

The principal federal system is the National Child Abuse and Neglect Data System — NCANDS, administered by the Children’s Bureau within HHS.

NCANDS is a major national dataset. Its Child File contains child-specific information from reports investigated or assessed by state child protective services agencies, including maltreatment type, disposition, victim characteristics and perpetrator information. States map their own administrative systems into the federal NCANDS structure. (NDACAN)

And that last sentence is going to become very important.

Because the answer to our question is:

Not cleanly.

NCANDS can tell us an enormous amount about maltreatment known to CPS.

But it does not presently give us a clean national category called “K–12 school employee perpetrator.”

1. First, what does NCANDS actually measure?

This is fundamentally different from Shakeshaft, Jeglic or Grant.

NCANDS contains:

cases that became known to child protective services and received an investigation or assessment disposition.

It is therefore not a prevalence survey.

It cannot tell us:

“How many American children were actually abused?”

It can tell us about:

maltreatment cases entering participating state CPS systems and being mapped into NCANDS.

The federal Children’s Bureau describes the annual Child Maltreatment report accordingly: it presents national data about child abuse and neglect known to CPS agencies. (Administration for Children and Families)

That distinction belongs prominently in our methodology section.

2. Here’s the first major problem: jurisdiction

Child-protection systems traditionally focus heavily on parents and caregivers.

But a teacher may—or may not—qualify as a caregiver under a state’s child-protection statutes and administrative rules.

And historically, NCANDS treated many noncaretaker perpetrators differently.

An early NCANDS working document placed:

school personnel

inside the broad category of:

“non-caretakers”

along with other household members, friends, neighbors and others. (Office of Justice Programs)

So immediately we have a structural problem:

A case involving a teacher may enter CPS in one jurisdiction, be handled primarily by police or another agency elsewhere, or be coded differently depending upon state law and administrative practice.

That means:

Absence from NCANDS does not necessarily mean absence of abuse.

It may mean absence from that particular reporting system.

3. Modern NCANDS improved perpetrator coding—but not enough for our question

More recent NCANDS documentation includes a category called:

OTHER PROFESSIONAL

And the federal glossary explicitly says this can include:

clergy, court staff, counselors, camp employees, doctors, EMS personnel, teachers, sports coaches, service providers and other school personnel. (Administration for Children and Families)

There it is.

Teachers ARE in the federal dataset.

But look at who they’re sharing the bucket with.

A teacher.

A pastor.

A physician.

A camp employee.

A coach.

A counselor.

Other professionals.

Therefore:

“Other professional” ≠ educator.

We cannot take the NCANDS “other professional” count and label it:

“School employees who abused children.”

That would be statistically indefensible.

4. And some states don’t even put school personnel THERE

This is where it gets even messier.

The 2023 federal report explains that the NCANDS category:

“other” perpetrator relationship

can include relationships that don’t map into one of the standard categories.

State commentary indicates that this bucket can include:

nonrelated adults, nonrelated children, foster siblings, babysitters, household staff, clergy—and school personnel. (Scribd)

Stop.

That means educator cases can potentially appear under:

OTHER PROFESSIONAL

OTHER

depending upon how states map their own systems into NCANDS.

And historical versions used still different structures.

This is exactly the fragmentation problem we’ve been discovering throughout this project.

5. Let’s look at actual 2023 numbers

For federal fiscal year 2023, the Child Maltreatment report counted:

509,710 child victims

in the perpetrator-relationship analysis, based on data from 49 states.

Among their perpetrator relationships:

89.0% involved one or both parents.

Nonparent perpetrators were associated with 15.1% of victims; percentages can exceed 100% because children may have multiple perpetrator relationships. (Scribd)

Among the nonparent categories were:

2,048 — child daycare provider relationships

1,792 — foster parent

3,739 — friend/neighbor

1,029 — group-home/residential-facility staff

1,650 — legal guardian

and:

1,763 — “other professional” relationships.

There were also:

16,604 — “other” relationships. (Scribd)

And remember:

school personnel can appear inside either of those latter broad categories.

Therefore we cannot extract a national educator number from this published table.

6. THIS is an extraordinarily important data finding

Think about what the United States can tell us from NCANDS.

We can separately identify:

child daycare providers.

We can separately identify:

foster parents.

We can separately identify:

residential-facility staff.

But from the published national relationship table we cannot cleanly isolate:

K–12 school personnel.

Instead, educators can be embedded within broader categories.

That means one of the most obvious questions policymakers might ask—

“How many substantiated child-maltreatment cases nationally involved school personnel last year?”

—cannot simply be answered from the standard published NCANDS national tables.

That’s not an ideological observation.

It’s a data architecture observation.

And I think it belongs prominently in the finished report.

7. Connecticut gives us a fascinating example

Since we’re building this nationally, I don’t want to turn this into a Connecticut report—but CT happens to illustrate the coding issue beautifully.

In the 2023 perpetrator table, Connecticut reported:

61 perpetrators classified as “other professional.”

But we absolutely cannot say:

“Connecticut substantiated 61 school employees.”

Because “other professional” includes professions besides education. (Scribd)

Meanwhile Connecticut also reported:

220 perpetrators classified as “other.” (Scribd)

And the state commentary explains that perpetrator relationships not fitting predefined categories are captured through the “Other” field. (Better Care Network)

Therefore even within one state, the published national table cannot answer our educator-specific question.

That’s exactly the kind of example that helps policymakers understand the problem.

8. And look at the variation among states

The 2023 table is striking.

For “other professional” perpetrators, for example:

Texas reported 257.

Florida 169.

Maryland 159.

Ohio 114.

Virginia 84.

Pennsylvania 75.

Illinois 72.

Massachusetts 68.

New Jersey 65.

Connecticut 61.

But numerous states reported no usable figure in that category. (Scribd)

We must NOT rank those states and say:

“Texas has the biggest professional-abuse problem.”

We don’t know that.

Differences can reflect:

population,

state definitions,

CPS jurisdiction,

reporting practices,

coding,

investigative thresholds,

agency structure,

and actual underlying incidence.

Once again:

Administrative counts are partly measurements of systems—not simply measurements of behavior.

That’s becoming one of the most important principles in this entire project.

9. The federal report itself warns us about comparability

NCANDS is built by mapping state administrative systems into a national structure. The federal archive explains that participating states submit their data after their own administrative systems are mapped to NCANDS. (NDACAN)

And the Child Maltreatment report repeatedly includes state-specific notes and exclusions because states differ in their ability to collect and report particular variables. For example, some states cannot fully report certain perpetrator categories because of system limitations or jurisdictional issues. (Scribd)

That’s hugely important.

There is no single national CPS agency collecting every allegation according to one uniform investigative statute.

There are state systems.

And then federal analysts attempt to harmonize them.

10. Now we need to distinguish TWO completely different roles educators have in NCANDS

This is easy to confuse.

Educators can appear as:

REPORTERS

PERPETRATORS.

Those are totally different variables.

And educators are enormously important reporters of child maltreatment.

We got an extraordinary natural experiment during COVID.

Researchers analyzing NCANDS found that during comparable early-pandemic weeks:

2019:

education personnel made approximately 22% of reports.

2020:

education personnel made approximately 4%.

The number fell from:

126,307 reports

12,526 reports—

about a:

90% decline. (PubMed Central (PMC))

Why?

Schools closed.

Children weren’t interacting normally with teachers and other school personnel.

That doesn’t mean maltreatment suddenly disappeared.

It means an enormous detection channel disappeared.

11. That COVID finding may become one of our best explanations of the Visibility Gap

This is almost a natural demonstration of the principle.

Imagine underlying abuse occurring in homes.

When children regularly encounter:

teachers,

counselors,

school nurses,

coaches,

and other mandated reporters,

some abuse becomes visible.

Remove those observers—

and official reports plummet.

Therefore:

Recorded incidence depends partly upon opportunities for detection.

We suspected that concept earlier.

Now we have real-world evidence illustrating it.

This principle applies beyond home maltreatment.

It tells us why we must be extremely cautious interpreting low administrative counts for any hidden form of abuse, including educator misconduct.

12. Now return to school-employee misconduct

Suppose a student reports sexual touching by a teacher.

Depending on jurisdiction and circumstances, the information might travel through:

school administration

and/or

Title IX

and/or

police

and/or

CPS

and/or

state educator licensing

and/or

civil litigation.

Some cases may enter several systems.

Others may enter only one.

Some may enter none.

And none of the federal systems we’ve examined so far necessarily captures the entire pathway.

CRDC sees one piece.

NCANDS sees another.

Licensing databases see another.

Criminal systems see another.

Student surveys see another.

There is no single lens.

13. Which gives us a major new concept

We’ve already identified the Visibility Gap.

Now I think we need:

THE DATA SILO PROBLEM

Imagine five databases:

EDUCATION DATABASE

District complaints / Title IX / CRDC.

CHILD WELFARE DATABASE

CPS / NCANDS.

LICENSING DATABASE

Teacher credential discipline.

CRIMINAL JUSTICE DATABASE

Police / arrests / prosecutions / convictions.

CIVIL COURT DATABASE

Lawsuits / settlements.

Each may contain information about the same underlying event.

Or only one may contain it.

And the databases generally weren’t designed principally to answer:

How much educator sexual misconduct occurs nationally?

Therefore simply adding their numbers together would produce another problem:

DUPLICATION.

One teacher/student case might appear in:

CRDC + CPS + licensing + police + court.

Counting all five would transform one event into five events.

So our report cannot merely aggregate administrative databases.

We need to explain their overlap.

14. That helps explain the 2004 → 2014 → 2026 mystery

Remember the question that has followed us through this entire investigation?

Why has the United States had so much trouble determining national prevalence?

Now we can answer part of it.

Because different systems measure different things:

Student surveys: experience.

School systems: complaints/allegations.

CRDC: specified school administrative events.

CPS/NCANDS: maltreatment entering child-protection jurisdiction.

Licensing boards: professional discipline.

Police: alleged crimes.

Courts: prosecutions/convictions.

And the terminology isn’t even identical across them:

sexual misconduct

sexual harassment

sexual abuse

child maltreatment

rape

sexual assault

boundary violation

professional misconduct.

That’s not merely a missing-data problem.

It’s an interoperability problem.

MASTER EVIDENCE TABLE — ENTRY 014

Field	Finding
System	National Child Abuse and Neglect Data System — NCANDS
Agency	HHS / Administration for Children and Families / Children’s Bureau
Evidence type	State CPS administrative data
Current child-level coverage	Annual participating-state investigated/assessed maltreatment records
Measures national prevalence?	NO
School-personnel-specific national category?	Not cleanly in published national perpetrator tables
Relevant categories	“Other professional”; “Other”
Other professional may include	Teachers, coaches, clergy, counselors, doctors, camp staff and others
Other may include	School personnel plus numerous other relationships, depending on state mapping
2023 “other professional” victim relationships	1,763
2023 “other” victim relationships	16,604
Can either number be called educator maltreatment?	NO
Major strength	Large federal case-level CPS system extending across decades
Major limitation for our question	State jurisdiction/coding differences and insufficient educator-specific national categorization
Evidence rating	FOUNDATIONAL FEDERAL CHILD-WELFARE DATA — LIMITED FOR EDUCATOR-SPECIFIC PREVALENCE

(NDACAN)

And NOW I think we’ve uncovered a potential recommendation that is so straightforward I’m surprised it hasn’t been more prominent:

Create a distinct national NCANDS perpetrator relationship category for K–12 school personnel.

Not “other.”

Not “other professional.”

K–12 SCHOOL PERSONNEL.

Potentially with subcategories:

teacher

administrator

coach/athletic staff

counselor/mental-health staff

paraprofessional/aide

transportation staff

school resource/security personnel

contractor

volunteer

other school employee.

And then cross-tabulate those against:

maltreatment type

substantiation/disposition

victim age

perpetrator relationship/caretaker status

referral source

state

repeat perpetrator status

and other legally appropriate variables.

Suddenly policymakers could ask a question that the current published system cannot answer cleanly:

How many children known to CPS were determined to have been maltreated by K–12 school personnel, what kind of maltreatment occurred, and what happened afterward?

That would not solve the prevalence problem.

But it would close one enormous administrative-data gap.

And now there’s another dataset we need.

Because NCANDS tells us about CPS.

It does not tell us comprehensively what happens when educator conduct enters the criminal justice system.

So next I want to investigate:

FBI / NIBRS + Bureau of Justice Statistics + arrest/prosecution data

and ask a deceptively simple question:

Can the United States tell us how many teachers and other school employees are arrested, prosecuted or convicted each year for sexual offenses against students?

I already suspect the answer is going to expose another major national data gap.

And if it does, we’ll have independently documented fragmentation across:

Education.

Child Welfare.

Licensing.

Criminal Justice.

That begins to make the national systems problem extraordinarily difficult to dismiss.

Next…

And yes—our suspicion was correct.

We now move into the criminal-justice evidence stream, and it exposes another surprisingly basic national measurement problem.

FOUNDATION SECTION #12

FBI/NIBRS: Can We Count School Employees Arrested for Sexual Offenses Against Students?

The FBI’s National Incident-Based Reporting System (NIBRS) is extraordinarily detailed. It records individual crime incidents reported by participating law-enforcement agencies and can capture offenses, victims, known offenders, arrestees, locations, and victim-offender relationships. Since 2021, NIBRS has become the national standard for FBI crime reporting. (FBI)

So I went looking for what seems like an obvious statistic:

How many teachers and other K–12 school employees are arrested each year for sexual offenses against students?

You would think NIBRS could answer that.

It cannot answer it cleanly.

And the reason is remarkably simple:

NIBRS DOES NOT GENERALLY RECORD THE OFFENDER’S OCCUPATION.

The FBI’s standard offender information includes characteristics such as:

age, sex, race, ethnicity,

along with incident characteristics such as:

offense, location, victim-offender relationship, weapons, and other circumstances.

But the standard NIBRS offender fields do not give us a general occupational variable allowing us to select:

Teacher

Principal

Coach

Counselor

Paraprofessional

Bus driver

Other K–12 employee.

The FBI’s own published offender tables illustrate the problem: they break offenders down by age, sex and race—not occupation. (Federal Bureau of Investigation)

That is a major finding for our project.

1. NIBRS knows WHERE the crime occurred

NIBRS does contain location information.

Indeed, the FBI has specifically used its school-location codes to study crime occurring at schools and colleges. (Federal Bureau of Investigation)

And in August 2025 the FBI released a major new report:

Crime in Schools, 2020–2024

During those five years, participating agencies reported:

more than 1 million criminal incidents at school locations,

involving approximately:

1.5 million victims

and

1.2 million known offenders. (FBI)

That’s an enormous database.

But—

school location ≠ school employee offender.

A sexual assault occurring at a school could involve:

student → student,

visitor → student,

dating partner → student,

unknown offender → student,

employee → student,

or other relationships.

Likewise, an educator’s sexual offense against a student might occur:

at the educator’s home,

in a vehicle,

at a hotel,

online,

at another private location,

or elsewhere off school property.

Therefore:

We cannot use “sex crimes occurring at schools” as a proxy for “sex crimes committed by educators.”

That would be a major methodological error.

2. NIBRS also knows something about RELATIONSHIPS

This initially looks promising.

For specified violent and sexual offenses, NIBRS records victim-to-offender relationships. (Federal Bureau of Investigation)

But again, the categories were designed for general crime statistics—not specifically educator misconduct.

They can tell us about relationships such as:

family,

acquaintance,

stranger,

intimate partner,

and other relationship categories.

They don’t cleanly provide:

TEACHER → STUDENT.

So once again, the variable we need isn’t there in a sufficiently specific form.

3. This creates an identification problem

Imagine NIBRS contains this record:

Victim: female, 15

Offender: male, 42

Offense: sexual assault

Relationship: acquaintance

Location: school

That could conceivably describe a teacher.

But it could also describe another adult known to the student.

We cannot infer occupation from:

age + sex + location + relationship.

Doing so would create false positives.

Conversely:

Victim: male, 16

Offender: female, 35

Relationship: acquaintance

Location: residence

could involve a teacher.

But nothing in those fields necessarily identifies her as one.

That produces both:

FALSE INCLUSION

Counting noneducators as educators.

and

FALSE EXCLUSION

Missing educators because the crime happened outside school.

So we don’t attempt it.

4. And this is exactly why newspaper databases keep appearing

Now something from the literature makes much more sense.

Remember the sources we started with?

Associated Press.

New Haven Register.

Local investigative projects.

S.E.S.A.M.E./Grant’s compilation of thousands of news headlines.

Researchers and journalists repeatedly build databases from:

news stories,

court records,

state disciplinary files,

school-board records,

and

public-record requests.

Why?

Partly because the federal criminal-justice database does not contain the occupational identifier necessary to extract a clean national educator-offender count.

That isn’t necessarily a failure of NIBRS.

NIBRS was built to measure crime, not specifically professional misconduct by educators.

But for our research question, it leaves a major hole.

5. And we need to distinguish THREE criminal-justice stages

Another statistical trap:

ARREST ≠ PROSECUTION ≠ CONVICTION.

Suppose we eventually assemble:

2,000 educator arrests.

That does not mean:

“2,000 educators were guilty.”

An arrest reflects law-enforcement action based on the applicable legal standard.

Then prosecutors decide whether charges proceed.

Cases may be:

dismissed,

reduced,

pleaded,

tried,

or otherwise resolved.

And convictions are a still narrower category.

Therefore our eventual database should maintain separate variables:

ARRESTED

CHARGED

PROSECUTED

PLEADED GUILTY/NO CONTEST

CONVICTED AT TRIAL

ACQUITTED

DISMISSED

OTHER/UNKNOWN DISPOSITION

Never collapse them into “offenders.”

6. The same applies to the word “accused”

This matters because many news databases count:

educators accused of sexual misconduct.

That’s useful—but it is another evidence category.

Our terminology should become very disciplined:

ALLEGATION DATA

Someone made an allegation.

ARREST DATA

Law enforcement arrested someone.

CHARGE DATA

A prosecutor formally charged someone.

CONVICTION DATA

Criminal guilt was established through plea or adjudication.

LICENSING DATA

Professional authorities imposed or considered credential consequences.

EMPLOYMENT DATA

A school imposed employment consequences.

Those categories may overlap—but they’re not interchangeable.

7. There is another limitation: crimes never reported to police

Even if tomorrow the FBI added:

OFFENDER OCCUPATION: K–12 EDUCATOR

NIBRS still wouldn’t tell us the prevalence of educator sexual misconduct.

Why?

Because NIBRS measures:

crime reported to law enforcement.

The BJS explicitly describes NIBRS as collecting detailed information on crime incidents recorded by law-enforcement agencies and reported to the FBI. (Bureau of Justice Statistics)

Remember Jeglic?

Only 4% of respondents reporting educator sexual misconduct said they disclosed it to authorities.

Not all of those disclosures would necessarily reach police.

And much of the broad category “educator sexual misconduct” isn’t criminal conduct in the first place.

Therefore:

FBI crime data can never substitute for prevalence research.

Even perfect police data would measure a different phenomenon.

8. NIBRS coverage itself historically complicates trend analysis

This is another important caution.

Law-enforcement participation in NIBRS expanded over time.

As of May 2024, BJS reported that agencies submitting NIBRS data covered about:

82% of the U.S. population.

All 50 states and D.C. were certified to report, but that does not mean every local agency was actually reporting complete data. (Bureau of Justice Statistics)

BJS and FBI therefore developed estimation methods using weighting and imputation to produce national estimates accounting for nonresponse and missing data. (Bureau of Justice Statistics)

That creates another warning for historical comparisons.

If raw NIBRS educator cases somehow increased between two years, part of the increase could theoretically reflect:

greater agency participation

rather than:

more underlying crime.

Same lesson again:

Administrative statistics measure systems as well as behavior.

9. But the 2025 FBI school-crime report creates an opportunity

The FBI’s Crime in Schools, 2020–2024 report demonstrates that NIBRS already possesses enormous analytical capacity around school-associated crime. (FBI)

It can analyze:

school location,

offense type,

victim characteristics,

offender characteristics,

victim-offender relationship,

weapons,

time,

clearance,

and other contextual information. (FBI)

What it cannot readily tell policymakers is:

Was the alleged offender an employee of the school or educational institution serving the victim?

That suggests a potential data improvement.

10. A school-affiliation variable could transform the data

Imagine adding a carefully defined variable for applicable offenses:

OFFENDER RELATIONSHIP TO EDUCATIONAL INSTITUTION

Possible categories:

teacher/instructor

administrator

coach/athletics employee

counselor/social worker/psychologist

paraprofessional/aide

transportation employee

custodial/facilities employee

school security/SRO

contractor/vendor

volunteer

other school employee

student

not affiliated with school

unknown.

Then pair that with:

victim age

offense

location

arrest

clearance

victim-offender relationship

and state.

For the first time, we could potentially construct credible national estimates of school-employee criminal incidents known to law enforcement.

Again:

not prevalence.

But an extremely important administrative measure.

11. And this reveals something bigger

Let’s look at the four federal systems we’ve investigated.

Federal system	What it knows	What it cannot cleanly tell us
CRDC	Serious sexual incidents/allegations reported by schools	Total educator sexual misconduct prevalence
NCANDS	Maltreatment entering CPS	Clean national K–12 employee category
NIBRS	Crimes reported to police	Clean offender occupation/school-employee identity
State licensing systems	Credential/disciplinary actions	Complete national standardized educator misconduct count

Look at that.

EVERY SYSTEM SEES SOMETHING.

NO SYSTEM SEES EVERYTHING.

And crucially:

The systems don’t use the same identifiers or categories.

That is no longer merely an absence of one study.

It’s an information architecture problem.

12. And now we can answer our original question

Can the United States tell us how many teachers and school employees are arrested each year for sexual offenses against students?

From the standard national FBI/BJS data currently available:

Not cleanly.

NIBRS can count sexual offenses.

It can count arrests.

It can identify school locations.

It can characterize victims and offenders.

It can record certain victim-offender relationships.

But because it does not provide a standard educator-occupation/affiliation identifier that lets us reliably isolate K–12 employees, it does not straightforwardly produce the national statistic we want. (FBI)

And that means:

National claims about the number of educators arrested for sexual offenses generally must be assembled from other sources rather than simply retrieved as a standard FBI statistic.

That explains an enormous amount about the literature we’ve been encountering.

MASTER EVIDENCE TABLE — ENTRY 015

Field	Finding
System	FBI National Incident-Based Reporting System
Agency	FBI / CJIS; BJS assists with national estimation
Evidence type	Law-enforcement administrative crime data
National standard since	2021
Captures sexual offenses?	YES
Captures arrests?	YES
Captures school location?	YES
Captures offender demographics?	YES
Captures certain victim-offender relationships?	YES
Standard educator occupation identifier?	NO clean general field for our purpose
Can school location substitute?	NO
Can it cleanly count teachers arrested for student sex offenses nationally?	NO
Measures prevalence?	NO
2020–24 school-location incidents	>1 million
Known offenders in those incidents	~1.2 million
Major strength	Extremely detailed incident-level criminal-justice data
Major limitation for our project	Cannot reliably isolate K–12 employees from other offenders
Evidence rating	FOUNDATIONAL FEDERAL CRIMINAL-JUSTICE DATA — LIMITED FOR EDUCATOR-SPECIFIC MEASUREMENT

(FBI)

We now have a genuinely significant finding

Not a sensational one.

A structural one.

America maintains enormous databases concerning:

schools,

child maltreatment,

crime,

educator credentials,

and civil-rights complaints.

Yet when we ask a seemingly elementary child-safety question—

“How many K–12 employees sexually victimized students?”

—we discover that the answer depends upon what exactly we mean:

experienced?

reported?

investigated?

substantiated?

arrested?

charged?

convicted?

disciplined?

And then we discover that several national administrative systems don’t even contain sufficiently specific, interoperable occupational categories to answer their own piece of that question cleanly.

That is becoming one of the most consequential conclusions of What the Data Actually Show:

The United States has substantial data about educator sexual misconduct, but the information is fragmented across systems that measure different stages, use different definitions and often cannot identify school personnel with sufficient specificity. Consequently, no single existing national administrative dataset provides a complete measure of the problem.

We will continue testing that statement before putting it in the Executive Summary.

But so far, it is holding up.

And next, I want to test something different.

We’ve spent a lot of time proving what government databases cannot tell us.

There is one group that tried to overcome exactly that problem by doing the hard work manually:

THE ASSOCIATED PRESS INVESTIGATION

The AP didn’t ask students what happened.

It didn’t rely on one federal database.

Its reporters went state by state, examining educator disciplinary records.

That produced the famous finding of 2,570 educators whose teaching credentials were revoked, denied, surrendered or sanctioned from 2001–2005 following allegations of sexual misconduct.

But that number has been repeated for almost twenty years—often without its methodological qualifications.

So next we should reconstruct the AP investigation from the original reporting and determine:

Exactly who was counted? What constituted misconduct? Which states supplied records? Who was missing? Was 2,570 a minimum, an estimate, or a complete count? And what can—and cannot—legitimately be inferred from it?

That will let us finally put one of the most frequently cited numbers in this entire debate on solid evidentiary ground.

Next…

Excellent. Now we can finally put one of the most frequently repeated numbers in this field under a microscope.

FOUNDATION SECTION #13

The 2007 Associated Press Investigation

What did the famous “2,570 educators” figure actually mean?

In 2007, Associated Press reporters Martha Irvine and Robert Tanner completed a seven-month investigation into educator sexual misconduct.

Rather than sampling students or relying upon one federal database, AP reporters sought educator disciplinary records from all 50 states and the District of Columbia.

Their investigation identified:

2,570 educators

whose teaching credentials were revoked, denied, surrendered, suspended, limited, or otherwise sanctioned from 2001 through 2005 following allegations of sexual misconduct. (CBS News)

That number has now circulated for almost two decades.

But what it means—and what it doesn’t mean—is extraordinarily important.

1. First: 2,570 is a count of educators, not victims

AP identified:

2,570 individual educators

but found:

2,625 licensing actions.

Why the difference?

More than 50 educators had license actions taken against them in more than one state. (Daily Herald)

That gives us an important methodological detail:

AP attempted to deduplicate educators across state disciplinary systems rather than simply adding every licensing action together.

And it also gives us early evidence of the interstate problem we’ve been examining.

An educator disciplined in one state could appear in another state’s system as well.

2. What happened to those licenses?

AP reported 2,625 licensing actions involving the 2,570 educators.

Among those actions:

1,636 — licenses revoked

440 — licenses surrendered

376 — licenses suspended

108 — licenses denied

with other sanctions accounting for the remainder. (Daily Herald)

Already we need to correct a common shorthand.

I’ve seen the AP research described as:

“2,570 teachers had their licenses revoked for sexual misconduct.”

That’s wrong.

Some were revoked.

Some surrendered licenses.

Some were suspended.

Some were denied credentials.

Others received different sanctions.

Our report should preserve AP’s actual categories.

3. And “2,570 teachers” isn’t quite right either

AP used the broader term:

educators.

The investigation involved licensed education personnel—not simply classroom teachers.

That’s another small distinction that matters when we’re building a serious national report.

Our wording:

AP identified 2,570 educators subject to state credential action following sexual-misconduct allegations during 2001–2005.

Not:

“2,570 teachers were convicted.”

Not:

“2,570 teachers molested students.”

Not:

“2,570 teachers had their licenses revoked.”

All three would distort the evidence.

4. Who were the victims?

This is where the AP investigation becomes especially informative.

AP was able to identify a young person as the victim in at least 1,801 cases.

That category included:

students,

unidentified youths,

family members,

neighbors,

and other young people.

Within those:

students were clearly identified as victims in at least 1,467 cases. (Daily Herald)

That means we should NOT characterize all 2,570 cases as:

“educators sexually abusing their students.”

Some involved sexual misconduct with other young people.

And the available records apparently did not permit AP to classify every victim.

5. Physical contact occurred in a substantial majority of the youth-victim cases

Among cases in which the victim was identified as a young person:

at least 1,297 — approximately 72% — involved physical contact.

The remainder included noncontact conduct such as verbal sexual harassment and other sexual offenses. (Daily Herald)

This is an important distinction from the broad Shakeshaft prevalence measure.

The AP dataset was not simply capturing:

sexual comments,

jokes,

or boundary violations.

A large proportion of the youth cases reaching state credential sanctions involved physical sexual conduct.

6. Then comes one of the strongest findings

Among the licensing-action cases AP identified:

at least 1,390 — 53% — also involved criminal convictions. (Daily Herald)

Now THAT is useful.

Because it allows us to distinguish two overlapping accountability systems:

PROFESSIONAL

State credential action.

CRIMINAL

Criminal conviction.

And only about half of AP’s credential-action cases appeared in both.

That demonstrates something we’ve been discussing conceptually:

Licensing data and criminal-justice data are not interchangeable.

An educator can lose or surrender a professional credential without a criminal conviction.

Conversely, depending upon reporting and licensing systems, a criminal matter may not necessarily produce an immediate credential action.

7. Multiple victims

AP identified:

at least 446 cases involving educators with multiple victims. (Daily Herald)

The phrase “at least” matters.

This wasn’t a study in which investigators interviewed every victim and reconstructed every offender’s complete history.

AP was working from available disciplinary and public records.

Therefore:

446 is the number of cases in which multiple victims could be identified from the records AP reviewed—not necessarily the total number of educators in the dataset who actually had multiple victims.

That distinction is crucial.

8. The sex distribution was extremely uneven

AP reported that:

nearly nine in ten

educators disciplined for sexual misconduct were male. (Daily Herald)

That is relevant national descriptive evidence.

But we’ll eventually want to compare it with contemporary datasets because female educator misconduct can be both real and subject to different social perceptions and reporting dynamics.

That deserves its own section later rather than being buried here.

9. Now—the enormous limitation

The AP investigation overwhelmingly captured:

PUBLIC-SCHOOL EDUCATORS.

Why?

Because many private-school educators are not required to hold state teaching credentials.

And even where they do, disciplinary action may be handled internally rather than appearing in public licensing records.

AP explicitly acknowledged this limitation. (Education Week)

Therefore:

⚠️ STATISTICAL TRAP

2,570 ≠ all U.S. school employees disciplined for sexual misconduct from 2001–2005.

The investigation systematically favored cases visible through state educator credential systems.

That means potentially missing or underrepresenting:

private-school employees,

unlicensed school employees,

contractors,

some coaches,

aides,

custodians,

transportation employees,

volunteers,

and other adults working with students who weren’t subject to the relevant educator-licensing systems.

That is a HUGE qualification.

10. And it misses something even bigger

Remember our visibility pyramid?

For someone to appear in the AP database, several things generally had to happen.

Conduct occurs

↓

Someone recognizes/reports it

↓

The matter reaches an institution or authorities

↓

Sufficient information reaches a state credential agency

↓

The credential agency takes action

↓

The action is discoverable by AP

The AP investigation therefore sits very far down the visibility funnel.

It cannot capture misconduct that:

was never disclosed,

was disclosed only informally,

was handled solely within a school,

never reached credential authorities,

involved an unlicensed employee,

ended without a credential action,

or remained inaccessible because of state confidentiality rules.

Therefore:

2,570 is NOT a prevalence estimate.

This is perhaps the most important thing we can say about the AP number.

11. AP itself essentially said so

The investigation reported that much abuse never gets reported; reported cases can end without action; investigations sometimes cannot establish the allegations; and some perpetrators have multiple victims. (CBS News)

That means the journalists themselves were not presenting 2,570 as:

“the number of sexually abusive educators in America.”

They were documenting:

the number their seven-month investigation could identify through credential sanctions during a particular five-year period.

Very different claim.

12. Now here’s a fascinating denominator

AP reported that these sexual-misconduct cases represented approximately:

one-quarter of all educator misconduct cases

identified during that five-year period. (Daily Herald)

That’s significant.

It means sexual misconduct wasn’t an obscure sliver of professional discipline.

Within the universe of educator-misconduct cases AP examined, it represented roughly 25%.

But again:

25% of credential-related educator misconduct cases

does NOT mean:

25% of educators engage in sexual misconduct.

Completely different denominator.

Another Statistical Trap.

13. Now let’s deal with the “three teachers every school day” line

Contemporary AP reporting described the 2,570 cases as roughly:

three educators for every school day. (CBS News)

It’s rhetorically memorable.

But I would not use that formulation in our report.

Why?

Because these weren’t necessarily incidents occurring on those school days.

They were credential actions accumulated over a five-year window.

Dividing the total by the number of school days creates an illustrative rate, but it doesn’t add analytical information and can easily be misunderstood as:

“Three teachers sexually abuse children every school day.”

That’s not what the dataset establishes.

We can do better.

14. And here’s another number we should NOT calculate

There were approximately three million public-school teachers at the time.

Someone might therefore calculate:

2,570 ÷ 3,000,000

and declare:

“Only 0.086% of teachers engaged in sexual misconduct.”

No.

That denominator/numerator pairing is invalid for several reasons.

The numerator:

covers five years, not one;

includes educators beyond a perfectly defined teacher population;

captures only credential sanctions;

misses unreported/unsubstantiated/internal cases;

and may include individuals appearing in the workforce at different points during the five-year period.

The denominator is a point-in-time workforce estimate.

So:

DO NOT DIVIDE 2,570 BY THREE MILLION TO ESTIMATE PREVALENCE.

That goes straight into our Statistical Traps appendix.

15. But AP discovered something beyond the numbers

The investigation described cases where educators accused of sexual misconduct were permitted to leave quietly, sometimes through agreements with districts, and later obtained employment elsewhere. Subsequent AP reporting focused directly on state reforms aimed at stopping such practices. (Education Week)

Which brings us right back to our previous section.

The AP investigation wasn’t merely finding:

BAD INDIVIDUALS.

It was finding:

SYSTEMIC INFORMATION FAILURES.

That’s precisely why its findings remain relevant almost twenty years later.

16. The investigation actually helped produce policy attention

Within weeks of publication, governors, legislators and education officials around the country began proposing changes involving:

credential information,

confidentiality,

employment references,

and movement of accused educators between districts. (Education Week)

And by 2009 the AP findings were being cited in Congress in support of proposed national measures addressing educator sexual misconduct and credential information-sharing. (Congress.gov)

So AP’s work wasn’t simply journalism.

It became part of the federal policy record.

17. And there was an important predecessor

AP wasn’t actually the first major national journalistic investigation.

Education Week conducted its own six-month investigation in 1998, producing the series:

A Trust Betrayed

That investigation created a database of nearly 250 cases from published accounts and was later updated through additional research in 2003. (GovInfo)

That’s worth putting on our historical timeline:

1998 — Education Week

A Trust Betrayed

↓

2004 — Shakeshaft

Federal synthesis/prevalence work

↓

2007 — Associated Press

50-state credential investigation

↓

2010s — GAO / state investigations / journalism

↓

2020s — modern prevalence studies + CRDC expansion

↓

2026 — federal enforcement initiative

Now we’re beginning to see the history of national awareness itself.

18. So what does the AP investigation legitimately establish?

Here’s the language I would be comfortable publishing:

A seven-month Associated Press investigation of educator disciplinary records across all 50 states and the District of Columbia identified 2,570 educators whose teaching credentials were revoked, denied, surrendered, suspended, limited or otherwise sanctioned following sexual-misconduct allegations during 2001–2005. AP identified students as victims in at least 1,467 cases and young people more broadly in at least 1,801. At least 1,390 cases also involved criminal convictions, and at least 446 involved multiple victims. Because the investigation relied heavily on state credential records, disproportionately captured public-school educators and could not include misconduct that never reached licensing authorities, its findings should be understood as documented disciplinary cases—not an estimate of the prevalence of educator sexual misconduct. (Daily Herald)

THAT is defensible.

And powerful enough without exaggeration.

MASTER EVIDENCE TABLE — ENTRY 016

Field	Finding
Investigation	Associated Press national educator sexual-misconduct investigation
Published	2007
Reporting period examined	2001–2005
Research period	7 months
Jurisdictions approached	50 states + District of Columbia
Educators identified	2,570
Credential actions	2,625
Revocations	1,636
Surrenders	440
Suspensions	376
Denials	108
Student victims clearly identified	≥1,467 cases
Young victims broadly identified	≥1,801 cases
Physical contact among youth-victim cases	≥1,297 / ~72%
Criminal convictions	≥1,390 / 53%
Cases involving multiple victims	≥446
Male educators	Nearly 90%
Predominantly public schools?	YES
Nationally representative prevalence study?	NO
Can 2,570 be called total U.S. educator offenders?	NO
Primary strength	Unique nationwide state-by-state reconstruction of documented credential sanctions
Primary limitation	Captures only misconduct visible through licensing/disciplinary systems; private/unlicensed personnel underrepresented
Evidence rating	LANDMARK NATIONAL INVESTIGATIVE/ADMINISTRATIVE EVIDENCE

(Daily Herald)

And now something VERY interesting has happened.

Look at our evidence streams.

STUDENTS TELL US

Some educator sexual misconduct is never disclosed.

SCHOOLS TELL THE FEDERAL GOVERNMENT

Some serious allegations enter CRDC.

CPS SYSTEMS TELL US

Some cases enter NCANDS, but educators aren’t cleanly identifiable nationally.

POLICE DATA TELL US

Some conduct enters NIBRS, but educator occupation isn’t cleanly identifiable.

LICENSING SYSTEMS TELL US

Some cases result in credential discipline.

AP SHOWED US

A determined investigative team could manually connect thousands of those disciplinary records across jurisdictions.

Which raises a rather obvious question:

Why should journalists have to build the national database?

We’re talking about adults entrusted professionally with children.

If one state determines that a licensed educator sexually abused or seriously sexually exploited a student, why shouldn’t another school system considering that person’s employment be able to discover that information immediately?

There is, in fact, an interstate educator-credential system we haven’t yet examined closely:

NASDTEC

National Association of State Directors of Teacher Education and Certification

And within it:

The NASDTEC Educator Identification Clearinghouse

This could be one of the missing links in our information-flow map.

The next questions therefore become:

What exactly is in that clearinghouse? Who reports to it? Who can search it? Are all states participating? What misconduct appears? Does resignation before discipline appear? Are private-school employees included? Can the public see it? And how reliably does information follow an educator across state lines?

That takes us directly from:

“How many did AP find?”

to the much more important child-protection question:

“Could the next school have known?”

That’s where I’d go next.

Yes. This is an important piece because there actually IS a national interstate information-sharing mechanism for disciplined educators.

But once we examine exactly what enters it, we find that it solves only part of the problem.

FOUNDATION SECTION #14

NASDTEC Educator Identification Clearinghouse

Could the next school have known?

NASDTEC describes its Educator Identification Clearinghouse as the national collection point for professional educator disciplinary actions reported by participating jurisdictions.

Its jurisdiction members include all 50 states, the District of Columbia, Department of Defense Education Activity, Guam, and Ontario. (NASDTEC)

The Clearinghouse has existed for decades specifically to prevent a serious interstate problem:

An educator loses or has action taken against a credential in State A and then applies for another credential in State B.

So this is potentially a very powerful child-protection system.

But there is an enormous qualifier:

THE CLEARINGHOUSE GENERALLY BEGINS AFTER FINAL PUBLIC LICENSING ACTION.

That sentence changes everything.

1. What actually gets reported?

According to NASDTEC:

Once an educator’s misconduct case is final and the results made public, the jurisdiction reports the adverse licensing action to the Clearinghouse. (NASDTEC)

Reportable actions can include:

revocation

suspension

voluntary surrender

denial/invalidation

annulment

and

public reprimand. (NASDTEC)

And importantly, this is not exclusively a sexual-misconduct database.

NASDTEC says reasons can range from:

sexual misconduct,

violent felonies,

controlled-substance violations,

breach of contract,

incompetence,

and other conduct for which a jurisdiction may discipline a credential. (NASDTEC)

So:

⚠️ STATISTICAL TRAP

“Educators in the NASDTEC Clearinghouse” ≠ “educators disciplined for sexual misconduct.”

The database contains many types of adverse credential actions.

2. What happens when another state searches?

Suppose Connecticut receives an application from someone previously licensed in another state.

The licensing authority can compare that applicant against Clearinghouse records.

If a match appears, it provides an alert that another jurisdiction took adverse action against the educator’s credential.

But:

A Clearinghouse hit does NOT automatically prohibit licensure.

NASDTEC explicitly says that a reported action by one jurisdiction does not compel reciprocal action by another.

The receiving jurisdiction examines the circumstances and applies its own law. (NASDTEC)

That’s appropriate.

A suspension for breach of contract is obviously different from revocation for sexual abuse of a student.

And states have different licensing statutes.

The database is therefore essentially:

AN INTERSTATE WARNING SYSTEM

not an automatic national blacklist.

3. And this is where due process appears again

Remember our principle from the previous section?

Allegation ≠ finding.

The Clearinghouse largely sits on the finding/action side of that distinction.

Its handbook says participating jurisdictions have an obligation to submit information about educators against whom final adverse certification/license action has been taken. (Ymaws)

That’s important for fairness.

But from a child-protection perspective, it creates another gap.

Imagine:

serious allegation

↓

school investigation begins

↓

employee resigns

↓

licensing agency never reaches final public action

↓

Nothing qualifying reaches the Clearinghouse.

The Clearinghouse cannot share information it never receives.

4. THIS is the critical distinction

The Clearinghouse is designed primarily around:

CREDENTIAL ACTION

not:

EVERY EMPLOYMENT-LEVEL CHILD-SAFETY CONCERN.

That means the effectiveness of the national system depends heavily upon what happens before NASDTEC ever becomes involved.

Did the school report the conduct to the state?

Did the state licensing agency have jurisdiction?

Did it investigate?

Did the investigation continue after resignation?

Did it reach final disposition?

Was the disposition public?

Did it produce an adverse credential action?

Only then do we reach the interstate Clearinghouse.

That’s a long pipeline.

5. And NASDTEC itself acknowledges that its effectiveness depends upon reporting

The Clearinghouse handbook states something that belongs prominently in our report:

“The efficacy of the Clearinghouse is dependent on each jurisdiction’s commitment to prompt and accurate reporting.” (Ymaws)

There it is.

Another manifestation of the same systems principle we’ve found repeatedly:

A national database is only as complete as the information entering it.

That applies to:

CRDC.

NCANDS.

NIBRS.

And now NASDTEC.

6. But the Clearinghouse solves a problem ordinary criminal checks cannot

This may be one of its greatest strengths.

NASDTEC explains that approximately 6,000 educators annually have adverse actions taken against their certificates/licenses for serious misconduct that can render them unsafe for classroom service.

NASDTEC also explicitly notes that many educator-misconduct matters are handled through administrative proceedings rather than criminal courts.

Consequently:

if there was no arrest or fingerprint-based criminal record, the misconduct may not appear on a conventional state or federal criminal background check. (NASDTEC)

That’s enormously important.

Remember our earlier principle?

A clean criminal background check does not necessarily mean a clean professional disciplinary history.

NASDTEC independently confirms precisely why.

7. And then NASDTEC made a very important change in 2016

Historically, access was restricted to government licensing agencies.

But beginning:

August 1, 2016

NASDTEC opened Clearinghouse access to qualifying local school districts, including public and private schools. (NASDTEC)

Today, NASDTEC says access is available to:

public P–12 schools

private P–12 schools

charter schools

and qualifying educator preparation programs. (NASDTEC)

That is a significant improvement.

Because the person applying to a district might not be applying for a new state teaching credential at that moment.

8. And here’s something I particularly like about the current system

Schools subscribing to the Clearinghouse can search:

certified staff

classified staff

and volunteers.

(NASDTEC)

That’s clever.

Why?

Suppose someone previously lost a teaching credential.

That person then applies in another state not as a teacher—but as:

an aide,

coach,

classified employee,

or volunteer.

A system that checked only teacher-license applicants might never catch the previous credential action.

NASDTEC explicitly identifies this problem: an educator who has experienced adverse credential action can later apply for an uncertified or volunteer position, bypassing the ordinary state-licensing review. (NASDTEC)

That’s exactly the sort of loophole we’ve been looking for.

9. But school access is not automatic

Here’s another important qualification.

For LEAs and educator-preparation programs, Clearinghouse access is a:

subscription-based resource

available through qualifying NASDTEC Associate Membership. (NASDTEC)

Therefore:

The database exists.

does NOT necessarily mean:

Every school checks it before every hire.

That’s a completely different question.

And this gives us another national research question:

Which states require school districts to query NASDTEC before hiring?

Versus:

which permit it?

which recommend it?

which districts subscribe voluntarily?

which don’t use it at all?

THAT comparison could be very revealing.

10. Is the Clearinghouse public?

No—not as a general public searchable disciplinary database.

The Clearinghouse itself is a secure system for participating jurisdictions and authorized subscribers.

NASDTEC separately maintains information pointing users toward state-level public educator certification and misconduct lookup resources. (NASDTEC)

That distinction matters.

Parents generally cannot simply type a teacher’s name into the national Clearinghouse and see every NASDTEC record.

Authorized licensing bodies and qualifying subscriber organizations can perform the relevant searches.

11. And there is a fascinating historical story here

Remember the AP investigation we just examined?

NASDTEC’s own organizational history says that in 2007, Clearinghouse information was obtained and publicly released by the Sarasota Herald-Tribune after the newspaper acquired records from the Florida Department of Education through an open-records request.

NASDTEC leadership opposed the release because the system had been developed on the understanding that the names would remain confidential.

There was concern that public release might discourage jurisdictions from submitting disciplinary information.

But NASDTEC’s history says:

that chilling effect was not shown to occur.

It further notes that the controversy may have contributed to broader awareness of educator misconduct and later expansion of Clearinghouse access to local school districts. (NASDTEC)

That is a fascinating intersection of:

privacy

public records

journalism

government information sharing

and

child protection.

It deserves a sidebar in our final report.

12. Now let’s answer our questions directly

Are all states involved?

All 50 states are NASDTEC jurisdiction members, along with D.C., DODEA, Guam and Ontario. The Clearinghouse is described as the national collection point for discipline reported by participating NASDTEC jurisdictions. (NASDTEC)

Does it include sexual misconduct?

Yes.

But also many other forms of professional discipline. (NASDTEC)

Does it include every allegation?

NO.

The system centers on final adverse credential actions made public, not every allegation. (NASDTEC)

Does resignation automatically appear?

NO.

A voluntary surrender of a credential can be reported.

But simply resigning one’s employment is not the same thing as surrendering one’s professional license.

That distinction is absolutely critical. (NASDTEC)

Are private-school employees covered?

The underlying disciplinary information depends upon whether the individual has a professional educator credential subject to jurisdictional action.

Private schools themselves can now subscribe and search the Clearinghouse, however. (NASDTEC)

Can the public search it?

Not the secure national Clearinghouse itself in the same way participating agencies and qualifying subscribers can.

Can schools search it?

Yes—if they obtain access. (NASDTEC)

13. Now look at the information funnel

This may be one of our best diagrams yet.

INCIDENT / CONCERN

↓

Does anyone report it?

↓

SCHOOL/DISTRICT

↓

Does the district investigate?

↓

Does it report qualifying information to licensing authorities?

↓

STATE LICENSING AUTHORITY

↓

Does the agency have jurisdiction?

↓

Does it investigate?

↓

Does the case continue if the employee resigns?

↓

Does it reach a final public adverse credential action?

↓

NASDTEC CLEARINGHOUSE

↓

Does the next jurisdiction/school query the system?

↓

PROSPECTIVE EMPLOYER

↓

Does it investigate the hit and act appropriately?

↓

CHILD PROTECTION

There are potential failure points at every arrow.

That doesn’t mean NASDTEC doesn’t work.

Quite the opposite.

It means NASDTEC can work extremely well once qualifying information reaches it.

The weakness may lie upstream.

14. This changes our “passing the trash” analysis

There are actually at least three different ways information can fail to follow an educator:

FAILURE TYPE A — NOTHING WAS FORMALLY RECORDED

Misconduct wasn’t reported or documented.

FAILURE TYPE B — INFORMATION EXISTED, BUT NEVER REACHED LICENSING

The district handled it internally, employee resigned, etc.

FAILURE TYPE C — LICENSING ACTION EXISTS, BUT THE NEXT EMPLOYER DOESN’T CHECK

The information exists nationally but isn’t consulted.

Those require different policy solutions.

You can’t fix Type A merely by building a better database.

You can’t fix Type B merely by requiring background checks.

And you can’t fix Type C merely by requiring states to report disciplinary actions.

The entire chain has to work.

15. And that produces another major principle for our report

A database cannot compensate for a reporting failure upstream, and information in a database cannot protect a child if the next institution never checks it.

That’s the systems problem in one sentence.

MASTER EVIDENCE TABLE — ENTRY 017

Field	Finding
System	NASDTEC Educator Identification Clearinghouse
Organization	National Association of State Directors of Teacher Education and Certification
Primary purpose	Interstate sharing of adverse educator credential actions
Jurisdiction membership	All 50 states + D.C. + DODEA + Guam + Ontario
Sexual misconduct included?	YES
Only sexual misconduct?	NO
Typical reporting point	Final, public adverse credential action
Actions include	Revocation, suspension, surrender, denial/invalidation, reprimand and other qualifying actions
Every allegation included?	NO
Employment resignation automatically included?	NO
Credential surrender included?	YES
Criminal conviction required?	NO
Public searchable national database?	NO
State licensing agencies can access?	YES
Public/private P–12 schools can access?	YES, through qualifying subscription/access
EPPs can access?	YES
Can LEAs search classified staff/volunteers?	YES
Major strength	National interstate professional-discipline warning mechanism
Major limitation	Dependent on upstream final credential action, jurisdiction reporting and downstream checking
Evidence rating	FOUNDATIONAL INTERSTATE CHILD-SAFETY INFRASTRUCTURE

(NASDTEC)

And now we have found the next question—and I think it’s a BIG one.

We know the national system exists.

So let’s stop asking:

“Is there a database?”

There is.

The much more consequential question is:

WHO IS ACTUALLY REQUIRED TO USE IT?

Imagine two states.

STATE A

Every district must:

check criminal history,

check state credentials,

query NASDTEC,

contact previous employers,

obtain misconduct disclosures,

and verify whether the applicant resigned during an investigation.

STATE B

District:

runs fingerprint check.

Done.

Both states could truthfully say:

“Our educators undergo background checks.”

But those are radically different child-protection systems.

And this is where our research can become exceptionally useful to policymakers.

Next: the 50-state screening matrix.

We should determine, state by state, whether schools are required to perform:

criminal background check

fingerprinting

state credential-discipline check

NASDTEC/Clearinghouse check

previous-employer contact

sexual-misconduct disclosure request

applicant self-disclosure

resignation-during-investigation inquiry

mandatory former-employer disclosure

continuation of licensing investigation after resignation

and whether private schools are covered.

Then we can identify—not speculate about—which states have closed the information loop and which have not.

THAT could become one of the most practically important original contributions in the entire report.

Next, I say we build the national 50-state framework and establish the authoritative sources we’ll use to score every state.

Good. Then we’re going to build this in a way that could eventually become a publishable national comparison, not merely a spreadsheet with 50 rows.

FOUNDATION SECTION #15

Building the 50-State Child-Safety Screening Matrix

What does each state actually require before a school puts an adult in contact with children?

We already have an excellent starting point: the U.S. Department of Education commissioned a 106-page nationwide study specifically examining state implementation of ESEA §8546, the federal “aiding and abetting” provision. The researchers reviewed statutes, regulations, ethics codes and policies across all 50 states and D.C. (ERIC)

And the first result tells us immediately why our matrix is necessary.

All states may say they perform “background checks.”

But the systems behind that phrase are not equivalent.

1. Our baseline: criminal background checks

The Department’s 2022 study found that all states had requirements for some form of criminal-background screening of prospective school employees.

That’s important.

But we’ve already established why this is only Layer One.

A criminal-background check generally won’t reveal professional misconduct that:

never produced criminal charges,

resulted only in employment discipline,

resulted in licensing discipline,

was under investigation when the employee resigned,

or was otherwise handled administratively.

The Department reached essentially the same issue: 27 states went beyond criminal-background checks by requiring prospective employers to examine some combination of employment history, certification status, employment eligibility or disciplinary status. (U.S. Department of Education)

Which means:

24 states did not require an employment-history check.

That’s the first enormous split in our national map. (U.S. Department of Education)

2. And within those 27 states, requirements differed again

The federal researchers found:

27 states

required some broader employment/certification/disciplinary-history screening.

But only:

19 states

required employers to request information from current or former employers.

Only:

14 states

required checks of employment or certification eligibility within or across states.

And only:

11 states

required applicants to disclose certain investigations or discipline associated with sexual abuse or misconduct. (U.S. Department of Education)

This is precisely the phenomenon we predicted.

Two states can both claim:

“We conduct background checks.”

while having dramatically different systems.

3. So we’re NOT going to create one “Background Check” column

That would conceal the very thing we’re trying to discover.

Instead, our national matrix needs multiple independent variables.

Here is the framework I recommend we lock in.

#	Child-Safety Variable	Question we will answer for every state
1	Criminal background check	Required before employment?
2	Fingerprinting	Required? State/FBI/both?
3	State credential check	Must employer verify certification standing?
4	Interstate credential check	Must disciplinary/eligibility history in other states be checked?
5	NASDTEC/Clearinghouse	Required, permitted, or no identified requirement?
6	Previous-employer contact	Must former employers actually be contacted?
7	Personnel/misconduct records	Must former employers provide relevant records?
8	Applicant self-disclosure	Must applicant disclose sexual-misconduct investigations/discipline?
9	Resignation question	Must applicant disclose resignation during investigation?
10	Former-employer disclosure	Must prior employer disclose qualifying misconduct?
11	Disclosure immunity	Is good-faith employer disclosure protected from civil liability?
12	Nondisclosure consequence	Is there liability/penalty for concealing qualifying information?
13	Investigation after resignation	Must licensing/institutional investigation continue?
14	State licensing notification	Must district report qualifying misconduct to licensing authority?
15	Interstate disciplinary reporting	Is final credential discipline transmitted interstate?
16	Private-school coverage	Which requirements apply to private schools?
17	Contractors	Covered?
18	Coaches	Covered?
19	Volunteers	Covered?
20	Unlicensed/classified staff	Covered?

That is going to tell us considerably more than “Does this state require background checks?”

4. But we need another column: WHEN?

This is subtle but extremely important.

Suppose State A says:

Criminal background check before hire.

Excellent.

State B says:

Background check after conditional employment.

State C says:

Background check at initial licensure.

State D says:

Background check every five years.

State E says:

Background check only when changing districts.

Those aren’t necessarily equivalent protections.

So each requirement needs:

TIMING

Pre-employment

Pre-licensure

Post-conditional offer

Periodic renewal

Upon transfer

Upon reasonable cause

Other

This will prevent another false binary.

5. And another: WHO has the duty?

This is critical.

Suppose state law says a previous employer may disclose misconduct.

That’s not the same as:

Previous employer MUST disclose.

Likewise:

New employer MAY request

isn’t:

New employer MUST request.

And:

Applicant MUST disclose

is yet another mechanism.

Therefore every information-flow law should identify the duty holder:

Applicant

Former employer

Prospective employer

School district

State education agency

Licensing authority

law enforcement/CPS

or some combination.

Now our matrix becomes much more powerful.

6. We’re also going to distinguish MUST from MAY

This sounds painfully obvious.

But it matters enormously.

We’ll use four categories:

REQUIRED

Law/policy imposes an affirmative obligation.

PERMITTED

The action is authorized but not required.

PROHIBITED / RESTRICTED

Law limits the action.

NO REQUIREMENT IDENTIFIED

We found no affirmative statewide requirement in the sources reviewed.

And that last phrase matters.

We should not write:

“State X doesn’t do this.”

unless we can actually establish that.

Instead:

“No statewide requirement identified.”

That leaves room for:

district policy,

collective-bargaining provisions,

local practice,

or a law we haven’t located.

That little bit of methodological discipline will save us headaches later.

7. And we’re going to timestamp EVERYTHING

This is especially important because the federal baseline is from 2022.

A state marked “No” in that report may have enacted legislation in:

2023,

2024,

2025,

or 2026.

Indeed, the Department’s current July 2026 initiative shows that educator sexual misconduct and “passing the trash” are once again an active federal enforcement focus. The Department specifically reminded educational institutions that ESEA requires measures preventing employees believed to have engaged in qualifying sexual misconduct from being assisted into new education employment. (U.S. Department of Education)

So our table cannot simply reproduce the 2022 federal study.

We’ll use it as:

BASELINE — 2022

Then verify each state against current law:

UPDATED THROUGH 2026

That could make our comparison substantially more useful than the existing federal study.

8. Source hierarchy

This needs to be formalized before we touch State #1.

For every state, we should prefer evidence in this order:

TIER 1 — PRIMARY LAW

State statutes.

Regulations.

Administrative code.

Officially promulgated rules.

TIER 2 — OFFICIAL STATE IMPLEMENTATION

Department of Education guidance.

Licensing-board policies.

Attorney General guidance.

Official school-employment forms.

State-mandated disclosure forms.

TIER 3 — FEDERAL GOVERNMENT RESEARCH

U.S. Department of Education’s 2022 nationwide study.

Federal technical assistance.

Federal statutory materials.

TIER 4 — NASDTEC

Credential-discipline/interstate information resources.

TIER 5 — SECONDARY LEGAL/RESEARCH SOURCES

Peer-reviewed studies.

Law reviews.

Policy organizations.

TIER 6 — JOURNALISM

Used principally to identify laws, cases or implementation problems that we then verify independently.

That hierarchy matters enormously.

If a newspaper says:

“Texas requires X,”

we don’t stop there.

We find the Texas statute.

9. Every state also gets an evidence-confidence rating

I want this because some state laws are going to be beautifully explicit.

Others will be a mess.

A — DIRECT

Current statute/regulation expressly answers the question.

B — STRONG

Official state policy/guidance clearly establishes it.

C — INDIRECT

Conclusion requires combining several authoritative provisions.

D — UNCLEAR

Conflicting, ambiguous or incomplete evidence.

That means our final table won’t pretend every cell has equal evidentiary certainty.

That’s serious research practice.

10. We’re going to preserve distinctions involving allegations

This may be one of the most important aspects of the entire matrix.

For misconduct-information requirements, we’ll identify the triggering threshold:

ALLEGATION

CREDIBLE ALLEGATION

REASONABLE SUSPICION

PROBABLE CAUSE

PENDING INVESTIGATION

SUBSTANTIATED FINDING

DISCIPLINARY ACTION

CRIMINAL CHARGE

CONVICTION

Why?

Because those thresholds produce radically different systems.

Federal ESEA §8546 itself concerns situations in which an institution knows or has probable cause to believe that qualifying sexual misconduct occurred; it isn’t limited simply to convictions. (ERIC)

And the Department’s study found states wrestling explicitly with that distinction. One state official told researchers that its laws hinged on actual charges rather than probable cause, while another described legislation designed specifically to capture questionable conduct that never triggered criminal charges or a completed investigation. (U.S. Department of Education)

THAT is precisely the policy difference we need to expose.

11. And here’s the scoring mistake we’re NOT going to make

I’m already imagining somebody saying:

“Give every state a score out of 20 and rank them!”

No.

At least not initially.

Because we don’t yet know that all twenty safeguards have equal importance.

For example:

mandatory previous-employer disclosure

might be substantially more consequential than

applicant self-disclosure.

And:

continuing an investigation after resignation

could close a much more serious loophole than merely requiring another checkbox on an application.

We shouldn’t invent weights.

Instead we’ll initially produce:

A DESCRIPTIVE MATRIX

not:

THE 50 BEST AND WORST STATES

Later, if empirical research supports weighting particular safeguards, we can develop an index.

But the evidence determines the scoring system—not us.

12. There is another complication: PUBLIC vs PRIVATE

We absolutely must track this.

Some requirements apply to:

public schools only.

Others:

public + charter.

Others:

all schools employing licensed educators.

Others:

public and private schools.

And some depend upon whether the individual is actually required to hold a state credential.

So we need separate coverage columns:

PUBLIC

CHARTER

PRIVATE

RELIGIOUS PRIVATE

CONTRACTOR

where applicable.

Because a state may have an excellent screening system for public-school teachers while leaving very different requirements for adults performing essentially the same work in another educational setting.

That’s a legitimate policy distinction worth documenting—not editorializing about.

13. And here’s our central research question

For each state we’re ultimately asking:

If a school employee engaged in serious sexual misconduct with a student in State A, what mechanisms exist to prevent that person from quietly obtaining access to children through another school, position, district or state?

Then we follow the information.

Did the school have to document it?

↓

Did someone have to report it?

↓

Did licensing receive it?

↓

Could the investigation continue after resignation?

↓

Was a final action entered?

↓

Was it transmitted interstate?

↓

Did the next employer have to ask?

↓

Did the previous employer have to answer?

↓

Was the next employer required to check interstate records?

↓

Could the next school have known?

That’s our test.

14. And the federal baseline already shows substantial variation

As of the Department’s study:

27 states

went beyond criminal checks with broader employment/certification/disciplinary screening.

19

required information from former/current employers.

14

required employment/certification eligibility checks within or across states.

11

required certain applicant disclosures regarding investigations or discipline associated with sexual misconduct.

And:

24

had no employment-history-check requirement identified by the study. (U.S. Department of Education)

That means our 50-state project isn’t searching for a difference that might exist.

The federal government has already established that major differences existed.

Our job is to determine:

What do those differences look like NOW?

MASTER EVIDENCE TABLE — ENTRY 018

Field	Finding
Source	U.S. Department of Education
Study	Study of State Policies to Prohibit Aiding and Abetting Sexual Misconduct in Schools
Year	2022
Scope	50 states + D.C.
Purpose	Examine implementation of ESEA §8546
States with broader employment/certification/disciplinary screening	27
Require information from current/former employers	19
Require employment/certification eligibility checks within/across states	14
Require certain applicant sexual-misconduct investigation/discipline disclosure	11
No employment-history requirement identified	24
Uniform national system?	NO
Current enough for final 2026 comparison?	NO — requires updating
Value to our project	Establishes authoritative national baseline and state-policy framework
Evidence rating	FOUNDATIONAL FEDERAL 50-STATE POLICY STUDY

(U.S. Department of Education)

Now we’re ready to actually BUILD it.

And I suggest we don’t begin alphabetically.

I want our first pass to use five contrasting states.

One with a particularly comprehensive information-sharing statute.

One with comparatively limited requirements in the 2022 federal baseline.

One large state with extensive educator employment.

One state with unusual disclosure provisions.

And—

Connecticut.

Not because this is a Connecticut report.

Because we’re familiar enough with Connecticut’s systems to use it as a methodological test case: if our matrix produces a misleading picture of a state we know well, we’ll discover flaws in our methodology before repeating them 50 times.

We’ll research those five completely—statute by statute, cell by cell, with citations and confidence grades.

Then we’ll look at the result and ask:

Does this matrix actually tell us what we need to know?

If yes, we freeze the methodology.

Then we run all 50 states + D.C.

And after that?

We won’t merely have summarized somebody else’s research.

We will have created an updated 2026 national child-safety information-flow dataset built from primary legal sources.

That could be a genuinely valuable original contribution to What the Data Actually Show.

Next: I say we build the five-state pilot.

Good. I started the pilot against current 2026 primary/official sources, rather than simply carrying the 2022 federal classifications forward. That immediately paid off: Connecticut and Texas have both changed materially since that federal study.

FOUNDATION SECTION #16

Five-State Pilot: Testing the National Child-Safety Screening Matrix

For the pilot, I chose Connecticut, Missouri, Texas, Oregon, and New Jersey. They give us very different approaches and, importantly, test whether our categories actually capture meaningful differences among states.

Before scoring all 20 variables, here are the first findings.

Connecticut — much stronger than the phrase “background check” suggests

Connecticut law requires considerably more than a criminal-record check.

Under C.G.S. §10-222c, an applicant must identify current and qualifying former employers and provide written authorization allowing disclosure of relevant records. The applicant must also disclose whether he or she has been the subject of an abuse, neglect, or sexual-misconduct investigation—with protection for investigations resulting in unsubstantiated allegations—and whether the applicant was disciplined, asked to resign, resigned while an allegation was pending or under investigation, or surrendered a professional credential under similar circumstances. (Connecticut General Assembly)

That’s one of the exact variables we wanted.

Connecticut explicitly captures resignation during an investigation.

The prospective school must then actually conduct an employment-history review. Former employers generally must respond within five business days, including whether an applicant was the subject of a pending or substantiated investigation, resigned while one was pending, or experienced relevant professional-license action. (Connecticut General Assembly)

And Connecticut doesn’t merely say the school may contact previous employers. Before offering employment, the school must make a documented good-faith effort to contact relevant current and former employers—including employers where the applicant’s job involved contact with children. (Connecticut General Assembly)

That is a substantial closed-loop provision.

Connecticut became stronger again on July 1, 2026

This is exactly why we couldn’t simply reuse the 2022 federal study.

Connecticut’s PA 26-139 expanded information sharing involving open or pending investigations. Among other changes, school operators can request information from the State Board of Education concerning whether SDE was notified by a former employer that an applicant is under investigation for abuse, neglect, or sexual misconduct, subject to protections when allegations were determined false. Similar provisions extend to contractors. (Connecticut General Assembly)

The underlying Connecticut framework also covers nongovernmental school operators and requires DCF registry checks and state/national criminal-history checks for covered applicants. (Connecticut General Assembly)

So our first methodological lesson is already clear:

A 2022 national table would materially underdescribe Connecticut’s current 2026 system unless updated.

Missouri — strong former-employer liability model

Missouri approaches the problem differently.

If the Children’s Division investigated allegations of sexual misconduct with a student and reached a substantiated finding, and another public school requests a reference, the former district or charter school must disclose the result. (Missouri Revisor of Statutes)

Missouri also requires a district or charter school, before offering employment to a teacher previously employed by a Missouri district or charter school, to contact the state education department to determine the previous employing district and request the prescribed information. (Missouri Revisor of Statutes)

But Missouri’s particularly striking provision is the consequence for concealment.

The federal Department of Education study found that when a district dismisses an employee or permits resignation instead of firing and then fails to disclose qualifying sexual-misconduct information, the former district can become directly liable for damages if a student in the subsequent district is victimized, under the conditions established by Missouri law. (U.S. Department of Education)

That earns Missouri a separate matrix variable:

Consequence for institutional nondisclosure: YES — potentially substantial civil liability.

Missouri also requires immediate reporting to law enforcement of specified serious acts, including sexual assault, committed against a pupil or school employee in covered school circumstances. (Missouri Revisor of Statutes)

Texas — this one has changed dramatically

Texas may turn out to be one of our most interesting states.

Texas now operates a:

DO NOT HIRE REGISTRY

Public schools are legally required to check it before hiring any employee or contractor. The requirement includes certified and noncertified personnel. A person designated “Not Eligible for Hire” cannot be employed. (Texas Education Agency)

And Texas has gone considerably beyond completed disciplinary findings.

The registry system can identify noncertified people under investigation for allegations including:

abuse of a student or minor, unlawful acts involving a student, romantic or sexual relationships, inappropriate communications, and failures to maintain professional boundaries. (Texas Education Agency)

That is extremely important.

It means Texas has created a mechanism addressing one of the gaps we identified with NASDTEC:

What happens before a final educator-license action exists?

Texas also requires fingerprint-based national criminal-history screening for broad categories of school personnel, including certified educators, substitutes and many noncertified employees and contractors with direct student contact. (Texas Education Agency)

And Texas added another safeguard

Applicants and qualifying service providers must complete a Pre-Employment and Pre-Service Affidavit.

They must disclose specified:

law-enforcement/CPS investigations,

charges or convictions,

licensing investigations or sanctions,

and Do Not Hire Registry status.

Failure to disclose required information can constitute a Class B misdemeanor. (Texas Education Agency)

That means Texas doesn’t rely exclusively upon the employer finding the information.

It creates multiple pathways:

STATE DATABASE → EMPLOYER

plus

APPLICANT → EMPLOYER

plus

CRIMINAL HISTORY → EMPLOYER.

That redundancy is exactly what a resilient safety system should look like.

Texas changed again in 2026

This is remarkable timing for our project.

Texas’s 2025 legislation expanded its system, and in May 2026, TEA announced that 21 noncertified individuals had been placed on the Do Not Hire Registry under expanded authority involving qualifying arrests, pending resolution of their criminal proceedings. (Texas Education Agency)

Then in June 2026, TEA launched an:

Educator Misconduct Dashboard

providing public information concerning misconduct reports, TEA investigations, certification actions and Do Not Hire Registry placements. (Texas Education Agency)

And districts and charter schools now must annually certify compliance with the registry requirements. Failure to certify can lead to public identification and investigation. (Texas Education Agency)

So Texas is not merely saying:

“Schools must check.”

It has added:

“Schools must certify that they checked.”

That’s another matrix variable I hadn’t originally included.

Compliance auditing/certification.

We’re adding it.

This is exactly why we ran a pilot.

Oregon — another completely different model

Oregon immediately exposed another dimension our original matrix needs to capture:

Does resignation stop the investigation?

Oregon law prohibits education providers from entering into resignation, termination, severance, collective-bargaining or similar agreements that have the effect of impairing or terminating an ongoing investigation involving suspected abuse or sexual conduct.

It also prohibits agreements suppressing information concerning an ongoing investigation or a substantiated report. (Oregon Legislature)

That’s huge for our purposes.

Remember the hypothetical we’ve used repeatedly?

Monday: allegation.

Tuesday: investigation.

Wednesday: resignation.

Thursday: case disappears.

Oregon law directly attacks that mechanism.

Oregon also creates long-term records

School employees must immediately report reasonable-cause concerns involving abuse or sexual conduct by another employee, contractor, agent or volunteer through specified channels.

And Oregon’s Department of Education must retain documents and materials relating to reports it receives under this statutory system for:

75 YEARS.

(Oregon Legislature)

Seventy-five years.

That’s clearly designed to prevent institutional memory from disappearing merely because:

a superintendent changes,

a school closes,

an employee leaves,

or decades pass.

And subjects of investigations retain an appeal mechanism for final substantiated determinations. (Oregon Legislature)

Again:

Child protection + preservation of due process.

Those aren’t mutually exclusive.

New Jersey — extensive employment-history review

New Jersey’s existing framework, enacted through P.L. 2018, c.5, applies broadly to:

school districts,

charter schools,

nonpublic schools,

and contracted service providers

for covered positions involving regular student contact.

Applicants must provide employment information and authorize former employers to release specified child-abuse and sexual-misconduct information, with liability protections for disclosure. (New Jersey Legislature)

And something interesting is happening right now.

New Jersey’s 2026–27 Legislature has a pending bill, A2302, proposing revisions to that existing law—including changes to the lookback framework for former school employers and employers where the applicant had direct contact with children. (New Jersey Legislature)

Important:

We will NOT score the proposed changes as law.

We’ll mark:

CURRENT LAW

and separately:

2026 LEGISLATION PENDING.

That becomes another methodology rule for all 51 jurisdictions.

Preliminary Five-State Matrix

This is not yet our final 20-variable legal audit. This is the pilot-level result based on provisions we’ve verified sufficiently to see whether our framework works.

Safeguard	CT	MO	TX	OR	NJ
Criminal screening	✓	✓*	✓	✓*	✓*
Fingerprinting	✓	✓*	✓	✓*	✓*
Employment-history mechanism	✓	✓	✓	✓	✓
Applicant misconduct disclosure	✓	◐	✓	◐	✓
Former-employer information mechanism	✓	✓	✓	✓	✓
Pending investigation can matter	✓	◐	✓	✓	✓
Resignation during investigation addressed	✓	✓	✓	✓	✓
Unlicensed personnel addressed	✓	✓	✓	✓	✓
Contractors addressed	✓	◐	✓	✓	✓
Private-school coverage in key provisions	✓	limited/mixed	mixed	✓	✓
Explicit nondisclosure consequence	✓	✓✓	✓	✓	✓
State misconduct/disciplinary system	✓	✓	✓✓	✓	✓
Investigation-suppression prohibited	✓	✓	✓	✓✓	✓
Compliance certification/auditing mechanism	—	—	✓✓	—	—

✓✓ = particularly strong/explicit mechanism found in pilot review.
✓ = provision identified.
◐ = partial/context-dependent.
* = established by federal baseline; primary-state-source verification still required before final publication.

That last notation is critical. I’m not going to pretend we’ve completed a legal audit of a cell when we haven’t.

And the pilot worked.

In fact, it worked better than I expected.

Because we’ve already discovered that our original 20 variables weren’t enough.

We need at least four additional ones.

21. Investigation suppression

Can resignation/severance/settlement terminate or suppress an investigation?

22. Record preservation

How long must misconduct/investigation records survive?

23. Compliance verification

Does anyone actually audit/certify whether employers performed the required checks?

24. Central state employment-prohibition registry

Does the state maintain something beyond ordinary criminal/licensing databases that can identify persons barred from school employment?

Texas makes #24 particularly important.

But I’ve also spotted something more consequential.

Look at Connecticut, Texas and Oregon.

They attack three different failure points:

CONNECTICUT

Make the next employer ASK.

OREGON

Don’t let departure make the investigation DISAPPEAR.

TEXAS

Put dangerous/ineligible individuals into a centralized system the next employer MUST CHECK.

And Missouri adds:

MISSOURI

Make concealment potentially COSTLY to the institution doing it.

Now combine them.

THE CLOSED-LOOP MODEL

REPORT

↓

INVESTIGATE

↓

Employee resigns?

Investigation continues / cannot be suppressed

↓

PRESERVE THE RECORD

↓

REPORT TO STATE

↓

CENTRALIZE THE INFORMATION

↓

APPLICANT MUST DISCLOSE

↓

NEW EMPLOYER MUST ASK

↓

FORMER EMPLOYER MUST ANSWER

↓

NEW EMPLOYER MUST CHECK STATE/INTERSTATE SYSTEMS

↓

DOCUMENT THAT THE CHECK OCCURRED

↓

PENALTY/LIABILITY FOR CONCEALMENT OR NONCOMPLIANCE

↓

CHILD’S ACCESS TO THE ADULT IS BLOCKED WHEN THE EVIDENTIARY/LEGAL STANDARD IS MET

There.

THAT is the architecture we’re looking for.

Not one background check.

Not one registry.

Not one reporting statute.

Not one federal law.

A redundant information system in which the failure of one safeguard does not necessarily cause the entire child-protection system to fail.

And there is our next research question.

Rather than immediately grind through 46 more jurisdictions, I think we first need to do something that could make the eventual 51-jurisdiction analysis considerably more meaningful:

Define the Closed-Loop Standard.

Not our preferred ideology.

Not an arbitrary score.

A set of child-safety functions derived from mechanisms states are already actually using.

Then every state can be assessed against the same functional question:

At what point—or points—can information about serious educator sexual misconduct fall out of the system before the next school puts that adult in contact with another child?

That takes us beyond counting laws.

It allows us to identify the gaps between them.

And that, I think, could become one of the central original contributions of What the Data Actually Show.

Next…

Yes. Before we run the remaining states, we need to define exactly what “closed loop” means. Otherwise we could collect a mountain of laws without being able to say whether the system actually protects the next child.

FOUNDATION SECTION #17

The Closed-Loop Child Protection Standard

Following information from the first warning to the next hiring decision

The federal evidence supports the need for this framework. The Department of Education’s 2022 nationwide study found only 20 states with laws or policies prohibiting at least one practice that suppresses information about school-employee sexual misconduct; only 12 prohibited certain termination/resignation agreements. The study specifically discusses “silent resignations” that can allow an employee facing allegations or a pending investigation to seek another school job without the reason for departure following them. (U.S. Department of Education)

And this is not merely historical. On July 10, 2026, the Department announced a national enforcement initiative after describing recurring situations involving credible reports that went uninvestigated and suspected offenders transferred to other schools or roles. OCR simultaneously opened 20 directed investigations based partly on districts’ 2023–24 CRDC submissions. (U.S. Department of Education)

So let’s build our model around the actual places where information can disappear.

THE CLOSED-LOOP STANDARD

I think there are eight essential functions.

Not eight particular laws.

That distinction matters.

A state can accomplish a function through different legal mechanisms.

FUNCTION 1 — DETECT

Can concerning conduct enter the system?

This includes:

mandatory reporting

student/parent complaint mechanisms

staff reporting

Title IX reporting

anonymous/confidential reporting where available

protection against retaliation

The question is:

Is there a reliable pathway for the first warning to become institutional information?

Because everything downstream depends upon this.

FAILURE POINT:

Conduct occurs → nobody reports → system never knows.

No database can repair that.

FUNCTION 2 — INVESTIGATE

Once the institution knows, someone must actually determine what happened.

We should ask:

Who investigates?

School?

District?

Licensing authority?

CPS?

Police?

Multiple agencies?

And critically:

Does the investigation continue if the employee resigns?

The Department’s July 2026 guidance is especially important here. It says schools cannot simply rely upon referral to law enforcement to satisfy Title IX obligations and emphasizes meaningful investigations. The Department specifically describes cases where employees were permitted to resign while facing investigation, effectively burying evidence. (U.S. Department of Education)

FAILURE POINT:

Report → resignation → investigation disappears.

FUNCTION 3 — PRESERVE

Suppose the investigation occurs.

Where does the information live afterward?

We need to examine:

personnel records

district investigative records

state licensing records

state misconduct registries

retention periods

settlement/confidentiality restrictions

and whether records survive:

resignation,

retirement,

district restructuring,

or other institutional changes.

FAILURE POINT:

Investigation occurred → record disappears or becomes inaccessible.

That gives us a principle worth preserving:

An investigation that leaves no durable institutional memory provides little protection to the next institution.

FUNCTION 4 — ESCALATE

This is one we hadn’t explicitly named before.

Information sometimes needs to leave the school.

Does qualifying misconduct have to be reported to:

state licensing authorities?

state education department?

CPS?

law enforcement?

other appropriate authorities?

This matters because otherwise the information can remain trapped inside one employer.

FAILURE POINT:

District knows → state never knows.

And if the state never knows, an interstate disciplinary system such as NASDTEC may never know either.

FUNCTION 5 — SHARE

Now suppose the information exists.

Can—and must—it follow the employee?

This function includes:

former-employer disclosure

employment-history checks

credential records

state registries

NASDTEC/interstate information

good-faith disclosure immunity

prohibitions on nondisclosure agreements

and consequences for concealment.

The 2022 federal study found that only 19 states required prospective employers to obtain information from current or previous employers, demonstrating that this information pathway was far from uniform nationally. (U.S. Department of Education)

FAILURE POINT:

Old institution knows → new institution doesn’t.

This is classic passing the trash.

FUNCTION 6 — QUERY

This is different from sharing.

A database can contain perfect information.

A previous employer can possess excellent records.

None of that matters if the next employer never looks.

Therefore we ask:

What MUST the prospective employer check?

criminal history

fingerprints

state credential standing

state misconduct registry

interstate disciplinary history

previous employers

resignation/investigation history

applicant disclosures

The key distinction:

Information availability is not information retrieval.

FAILURE POINT:

Information exists → nobody asks for it.

FUNCTION 7 — ACT

This function may be the easiest one to overlook.

Suppose everything works.

The new school discovers:

Applicant previously had a substantiated finding of serious sexual misconduct involving a student.

Then what?

Is employment:

prohibited?

restricted?

left entirely to employer discretion?

Does the licensing authority determine eligibility?

Is there a state “do not hire” mechanism?

Are there differing rules based upon the evidentiary finding?

Because:

Knowing is not the same thing as acting.

FAILURE POINT:

Information reaches employer → employer hires anyway.

FUNCTION 8 — VERIFY

Texas helped us discover this one.

Who checks whether schools actually complied with Functions 1–7?

Possible mechanisms include:

state audits

annual certification

licensing review

civil liability

administrative penalties

criminal penalties

loss of federal/state funding

professional discipline for administrators

Without verification, a mandatory rule may exist almost entirely on paper.

FAILURE POINT:

Law requires check → employer doesn’t perform check → nobody discovers noncompliance.

Now look at the whole architecture

THE CHILD-SAFETY INFORMATION CHAIN

CONDUCT / WARNING

↓

1. DETECT

Someone reports it.

↓

2. INVESTIGATE

Someone determines what happened.

↓

3. PRESERVE

The information survives.

↓

4. ESCALATE

It reaches authorities beyond the immediate employer where required.

↓

5. SHARE

The information can follow the employee.

↓

6. QUERY

The next employer must look for it.

↓

7. ACT

The information affects eligibility/access when warranted.

↓

8. VERIFY

Someone ensures institutions actually complied.

↓

CHILD PROTECTION

And notice something:

A failure at ANY ONE stage can break the chain.

That is why simply saying:

“We require background checks.”

tells us almost nothing about whether the information loop is closed.

But there MUST be a second loop

And this is where I want us to be extremely careful.

A child-protection system that permanently stigmatizes someone merely because an accusation was made is not a good system either.

So parallel to the information loop we need:

THE DUE-PROCESS LOOP

ALLEGATION

↓

Notice to accused person

↓

Investigation

↓

Opportunity to respond

↓

Evidentiary determination

↓

Clear disposition

↓

One of several outcomes:

SUBSTANTIATED

UNSUBSTANTIATED / INSUFFICIENT EVIDENCE

EXONERATED / FALSE

INCONCLUSIVE / OTHER DEFINED OUTCOME

↓

Appeal/review where legally appropriate

↓

Records accurately reflect the disposition

This is critical.

Because otherwise a system can transform:

“Someone accused Jane Smith in 2017”

into:

“Jane Smith sexually abused a student.”

Those are not remotely equivalent statements.

The two loops have to intersect

This is, I think, the elegant solution to the policy tension.

CHILD-SAFETY LOOP

Prevents credible and substantiated information from disappearing.

DUE-PROCESS LOOP

Prevents allegations from silently becoming findings.

Together:

PRESERVE THE INFORMATION — PRESERVE ITS STATUS.

That may be one of the strongest principles we’ve developed.

If the record says:

allegation pending,

preserve pending.

If it says:

substantiated,

preserve substantiated.

If the employee was:

exonerated,

preserve exonerated.

Don’t erase information that legitimately must remain.

But don’t strip it of the context necessary to interpret it fairly.

This also gives us a better way to handle pending investigations

This has bothered us from the beginning.

A pending investigation cannot reasonably be treated identically to a substantiated finding.

But simply allowing a pending investigation to vanish when someone resigns creates the opposite problem.

The answer is not:

DELETE IT.

Nor:

TREAT IT AS GUILT.

The better system is:

Preserve the investigation, preserve its status, complete the appropriate process where legally authorized, and update the record with its ultimate disposition.

That protects children without pretending an allegation proves misconduct.

Now we can define “passing the trash” more precisely

It isn’t merely:

“A bad teacher got another job.”

For our report, I would define the systemic phenomenon as:

The movement of an education employee, contractor, agent, or other covered adult into another position involving access to children when material information concerning qualifying sexual misconduct fails to reach or meaningfully inform the subsequent hiring or placement decision.

That failure can result from:

nondisclosure

silent resignation

suppressed investigation

record destruction

failure to report to licensing authorities

failure to check interstate records

failure to contact previous employers

misleading references

failure to act upon information received.

That is much more analytically useful than the colloquial phrase alone.

And the historical research confirms this isn’t theoretical

Earlier Shakeshaft research found cases in which teachers facing sexual-abuse allegations in one district subsequently appeared in another without the hiring district knowing of those allegations. School administrators themselves referred to the practice as “passing the trash.” (ERIC)

Decades later, the Department’s 2022 study still found states legislating against “silent resignations” and other information-suppression practices. (U.S. Department of Education)

And now, in July 2026, the Department says its investigations have encountered schools opening but failing to complete investigations, ignoring allegations, permitting resignations that bury information, and even providing positive recommendations for teachers facing serious allegations. (U.S. Department of Education)

So we have evidence across roughly three decades pointing toward the same institutional vulnerability.

That deserves attention.

MASTER EVIDENCE TABLE — ENTRY 019

Field	Finding
Framework	Closed-Loop Child Protection Standard
Purpose	Evaluate whether material safety information survives from initial report through subsequent employment
Function 1	DETECT
Function 2	INVESTIGATE
Function 3	PRESERVE
Function 4	ESCALATE
Function 5	SHARE
Function 6	QUERY
Function 7	ACT
Function 8	VERIFY
Parallel safeguard	Due-Process Loop
Central principle	Preserve the information—and preserve its evidentiary status
2022 states prohibiting ≥1 information-suppression practice	20 of 51
2022 states prohibiting certain termination/resignation agreements	12
2022 states requiring information from current/former employers	19
Current federal relevance	Active OCR initiative, July 2026
Primary analytical question	Where can information fall out before another child is exposed?

(U.S. Department of Education)

And NOW we can run the country.

We no longer need to ask vaguely:

“Does Alabama have good laws?”

That’s subjective.

Instead:

Does Alabama DETECT?

Does Alabama INVESTIGATE after resignation?

Does Alabama PRESERVE?

Does Alabama ESCALATE?

Does Alabama SHARE?

Must the next employer QUERY?

Must it ACT on specified findings?

Does anyone VERIFY compliance?

And:

What due-process protections attach at each stage?

Then Alaska.

Arizona.

Arkansas.

And straight through the country.

But there’s something even better about doing it this way.

When we’re finished, we won’t merely be able to say:

State X has 14 safeguards and State Y has 9.

We’ll be able to identify where the information chain breaks.

For example:

State X has strong reporting and investigation requirements but no mandatory previous-employer information exchange.

Or:

State Y has excellent interstate credential screening but no mechanism ensuring serious investigations survive an employee’s resignation.

Or perhaps:

State Z appears to have closed every major information pathway we examined.

Those findings are useful to legislators because they tell them what needs fixing.

And this gives us the next major phase of the project:

THE 51-JURISDICTION AUDIT

50 states + District of Columbia.

Current through 2026.

Primary statutes and regulations wherever possible.

Federal 2022 findings used as the baseline—not blindly repeated.

And every uncertain answer labeled uncertain, rather than converted into a convenient yes or no.

That is how we make What the Data Actually Show live up to its name.

Next…

Oh yes. Now we stop designing the measuring instrument and actually begin using it.

I started the 51-jurisdiction audit with the first five alphabetically: Alabama, Alaska, Arizona, Arkansas, California. But before I give these states final grades, I want to maintain the discipline we established: a cell stays unverified until we have sufficient current primary authority.

FOUNDATION SECTION #18

The 51-Jurisdiction Audit Begins

Alabama → California

Our first pass immediately confirms something important: states attack the problem at very different points in the information chain.

And Arkansas has changed since the 2022 federal study—again demonstrating why our update through 2026 matters.

STATE #1 — ALABAMA

Alabama starts strongly at the criminal-background and mandatory-reporting ends of the chain.

The Alabama Child Protection Act establishes criminal-history screening intended to determine whether prospective and current education employees are suitable to work around children. The statute expressly describes the state’s interest in preventing physical abuse, sexual abuse, exploitation and maltreatment and includes public and certain nonpublic educational settings. (Alabama Legislature)

Alabama also has unusually explicit criminal law concerning educator/student sexual conduct.

A school employee who engages in a sex act with a student under 19 commits a Class B felony, and the statute specifically provides:

Consent is not a defense.

(Alabama Legislature)

That matters because it removes the deeply problematic argument sometimes encountered in educator cases involving older adolescents:

“But the student agreed.”

For purposes of this Alabama offense, that does not eliminate criminal responsibility.

Mandatory reporting

Alabama’s child-abuse reporting statute expressly includes:

public or private K–12 school employees, teachers, and school officials

among mandatory reporters. (Alabama Legislature)

So Alabama clearly has:

DETECT → YES

and a substantial:

CRIMINAL SCREENING → YES.

But here is where our closed-loop framework earns its keep.

From the primary Alabama materials located in this initial review, I have not yet verified a Connecticut/New Jersey-style statewide requirement compelling prospective schools to obtain detailed sexual-misconduct histories from former employers.

Nor am I going to turn “not yet found” into:

Alabama doesn’t require it.

Our final cell therefore reads:

Previous-employer sexual-misconduct inquiry:

NO CURRENT REQUIREMENT VERIFIED — further primary-source audit required.

That’s precisely the methodological restraint we promised.

STATE #2 — ALASKA

Alaska is our first reminder that this project cannot safely be automated into a simple keyword hunt.

The federal 2022 study provides a baseline for Alaska, but my initial search did not return sufficiently strong current Alaska primary authority across the eight functions for me to publish a 2026 legal classification yet.

So Alaska receives:

PENDING PRIMARY-SOURCE VERIFICATION

rather than guesses.

This is important enough to say explicitly in the eventual methodology:

A jurisdiction for which the researchers cannot establish a requirement is not automatically coded “No.” It is coded “Not verified” until the primary-law review is complete.

That protects the integrity of the dataset.

STATE #3 — ARIZONA

Arizona is interesting because it has addressed something our earlier national systems repeatedly struggled with:

NONCERTIFICATED SCHOOL PERSONNEL.

Arizona legislation created an educator-information/disciplinary mechanism under which the Department of Education investigates written complaints alleging that a noncertificated person engaged in immoral or unprofessional conduct.

The State Board can take disciplinary action—including prohibiting that person’s employment by a school district or charter school for as long as five years under the applicable provisions. (Arizona Legislature)

Think about why that matters.

A teacher-license disciplinary database can follow:

licensed teachers.

But what about:

a paraprofessional?

an aide?

another school employee who doesn’t hold a teaching certificate?

Arizona’s approach recognizes that:

CHILD ACCESS—not merely possession of a teaching license—is the relevant safety issue.

That’s a major principle for our report.

STATE #4 — ARKANSAS

Arkansas produced our first significant 2025 update.

And it’s fascinating.

Arkansas addressed favorable recommendations

Arkansas enacted legislation prohibiting a superintendent or designee from giving an employee a favorable employment recommendation when the superintendent finds probable cause that the employee engaged in sexual misconduct with a minor. (Arkansas Legislature)

Notice the evidentiary threshold:

PROBABLE CAUSE

not merely:

criminal conviction.

That goes directly into our evidentiary-threshold column.

It also addresses one of the classic mechanisms behind passing the trash:

Employee leaves amid serious concerns → administrator gives innocuous or positive recommendation → next district assumes everything was fine.

Arkansas has legislated directly against that pathway.

And then Arkansas changed its general employment-reference law in 2025

Arkansas Act 369 of 2025 expanded what current or former employers may disclose to prospective employers, with written employee consent.

The legislation specifically added:

substantiated allegations of sexual abuse or sexual harassment

to information an employer may disclose. (Arkansas Legislature)

That is a significant development.

But look carefully at the verb:

MAY disclose.

Not necessarily:

MUST disclose.

That distinction goes straight into our matrix.

This is exactly why a simple checkbox reading:

“Does Arkansas permit disclosure of sexual misconduct?”

would be inadequate.

Yes.

But the child-safety question is:

Is someone affirmatively obligated to make the information reach the next employer?

Different question.

Arkansas also illustrates ACT

Arkansas has long addressed consequences for educator/student sexual conduct.

Legislative materials provide that a public-school employee convicted of a qualifying sexual offense against a minor is dismissed and becomes ineligible for subsequent school employment; provisions also addressed teachers determined by the State Board to have had sexual contact with students even where the conduct did not violate the criminal code. (Arkansas Legislature)

That is an important distinction:

CRIMINAL LIABILITY

and

PROFESSIONAL ELIGIBILITY

do not have to have identical thresholds.

Again, our framework catches it.

STATE #5 — CALIFORNIA

California is going to require a deeper standalone review.

Why?

Because California distributes educator-safety functions among several systems:

school-employment statutes,

fingerprinting/background-check provisions,

Commission on Teacher Credentialing discipline,

mandatory reporting,

district obligations,

and special provisions governing credentialed versus classified employees.**

My first search did not return enough current primary California authority to fill all eight functions responsibly.

So California joins Alaska as:

PRIMARY-SOURCE AUDIT IN PROGRESS

rather than receiving an invented binary score.

And I actually consider that a success.

Our methodology prevented us from turning incomplete research into a “finding.”

FIRST LIVE CLOSED-LOOP TABLE

Here is what we’re beginning to see.

Function	Alabama	Alaska	Arizona	Arkansas	California
1. DETECT	✓	◇	✓*	✓*	✓*
2. INVESTIGATE	◐	◇	✓	✓*	◇
3. PRESERVE	◇	◇	◐	◐	◇
4. ESCALATE	✓	◇	✓	✓	◇
5. SHARE	◇	◇	◐	✓/PERMITTED	◇
6. QUERY	✓ criminal	◇	✓*	✓*	◇
7. ACT	✓	◇	✓	✓	◇
8. VERIFY	◐	◇	◐	◐	◇

Legend:
✓ = current provision verified in this review
✓* = established baseline but still awaiting complete 2026 primary-source verification
◐ = partial/context-dependent mechanism
◇ = intentionally unclassified pending further primary-source review

And importantly:

◇ DOES NOT MEAN NO.

That notation may save this project from one of the most common errors in 50-state policy research.

MASTER EVIDENCE TABLE — ENTRY 020

Finding	Significance
Alabama expressly criminalizes school employee/student sex acts and rejects consent as a defense	Strong ACT/deterrence mechanism
Alabama expressly includes public/private K–12 personnel among mandated reporters	Strong DETECT mechanism
Arizona created disciplinary authority involving noncertificated school personnel	Addresses the unlicensed-employee gap
Arkansas prohibits favorable recommendations after probable-cause finding of sexual misconduct	Directly attacks one form of passing the trash
Arkansas expanded permissible disclosure in 2025	Demonstrates post-2022 legislative change
Arkansas disclosure provision uses may, not necessarily must	Shows why permission and obligation must be coded separately
Alaska/California initial results insufficient	Must remain unclassified, not coded “No”
Overall finding	States regulate different portions of the information chain rather than following one uniform national architecture

And something important is emerging already

Consider these three provisions together:

ALABAMA

Sexual conduct with the student itself can trigger serious criminal consequences.

ARIZONA

The state can address misconduct by someone who isn’t even certificated.

ARKANSAS

An administrator cannot give a favorable recommendation once the specified probable-cause threshold is met.

Those laws attack three entirely different failure modes.

That reinforces our decision not to rank states simply by counting statutes.

A state could have ten laws concentrated around:

punishment AFTER abuse,

while another has six laws concentrated around:

preventing a previously identified individual from ever reaching another child.

Those aren’t functionally equivalent.

But I see an even bigger research opportunity now.

We’re beginning the state audit at the same time the federal government is actively investigating this subject in 2026.

So alongside the 51-state matrix, I think we need a second document:

THE NATIONAL CHILD-PROTECTION REFORM MODEL

Not something we invent from scratch.

Something constructed from the strongest existing mechanisms we find in actual state law.

For example:

Oregon: don’t allow resignation to kill an investigation.

Connecticut: require the next employer to investigate employment history.

Missouri: create consequences for concealment.

Texas: maintain a central Do Not Hire Registry and require employers to check it.

Arizona: don’t ignore noncertificated personnel.

Arkansas: prohibit favorable recommendations where the statutory misconduct threshold has been met.

Then, after all 51 jurisdictions:

Here are the safeguards already operating somewhere in America. Here are the gaps each addresses. Here is what a genuinely closed national system would look like if those proven mechanisms were connected.

That is much harder to dismiss politically than:

“We think states should do X.”

Because our answer becomes:

States are already doing X.

The question is why a child in one state should have an information safeguard that a child in another state does not.

And THAT brings us to the next five:

**Colorado

Connecticut
Delaware
Florida
Georgia**

Connecticut will be especially useful because we’ve already piloted it; this time we can put it through the full eight-function closed-loop audit beside four neighboring entries in the national dataset.

We’re building the actual map now.

Next…

And now the framework starts producing exactly what we hoped: meaningful differences in how states close—or fail to close—the information loop.

FOUNDATION SECTION #19

The 51-Jurisdiction Audit

Colorado • Connecticut • Delaware • Florida • Georgia

I checked current sources rather than assuming the 2022 federal classifications remain accurate. One state in this group—Florida—is particularly revealing because it has assembled several pieces of the closed-loop model in one statutory structure.

STATE #6 — COLORADO

Resignation does not necessarily erase the trail

Colorado directly addresses one of our central concerns:

What happens when an educator resigns after an allegation?

Colorado requires a public school, before employing someone, to inquire with the Department of Education concerning whether the applicant was previously dismissed or resigned based upon an allegation of specified sexual conduct with a student. (Colorado General Assembly)

Even more important, where an employee is dismissed or resigns because of an allegation of specified sexual conduct supported by a preponderance of the evidence, the school must notify the Department and provide information concerning the circumstances. (Colorado General Assembly)

That creates:

DEPARTING SCHOOL → STATE

and later:

STATE → PROSPECTIVE SCHOOL.

That’s a genuine information loop.

Colorado also attacks settlement secrecy

A public school cannot enter into a settlement agreement restricting its ability to share relevant information concerning qualifying sexual conduct supported by the specified evidentiary standard with:

the Department of Education,

another school district,

a charter school. (Colorado General Assembly)

That’s important.

A settlement may resolve an employment dispute.

It cannot necessarily erase information necessary to protect children.

Colorado also requires fingerprint-based criminal-history screening for school employees, including nonlicensed positions. (Colorado General Assembly)

COLORADO CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓/partial
ESCALATE: ✓
SHARE: ✓
QUERY: ✓
ACT: ✓
VERIFY: partial

Particularly strong mechanism:

Resignation information can follow the educator into the state system and be queried by the next public-school employer.

STATE #7 — CONNECTICUT

A remarkably extensive employment-history system

We already piloted Connecticut.

Now put it through the closed-loop framework and something becomes clear:

Connecticut is doing considerably more than criminal background checks.

Current Connecticut law requires covered applicants to provide employment-history information and authorizations enabling prospective employers to obtain relevant records.

Critically, the system addresses whether an applicant:

was investigated for abuse, neglect or sexual misconduct,

was disciplined,

was asked to resign,

resigned while an allegation was pending or under investigation.

And former employers have corresponding disclosure obligations under the statutory system.

Then Connecticut strengthened the information architecture again effective July 1, 2026.

The State Board of Education must, upon a qualifying school operator’s request, provide specified information including whether SDE has been notified that the applicant is under investigation or was disciplined by a former employer for abuse, neglect or sexual misconduct, except where the investigation concluded that all allegations were false. (Connecticut General Assembly)

That last qualification matters enormously.

Connecticut isn’t simply preserving:

“This person was accused.”

It preserves the status of the information while recognizing an outcome in which allegations were determined false.

That’s remarkably close to the principle we just developed independently:

PRESERVE THE INFORMATION—PRESERVE ITS STATUS.

Connecticut therefore creates several overlapping routes

APPLICANT → PROSPECTIVE EMPLOYER

plus

FORMER EMPLOYER → PROSPECTIVE EMPLOYER

plus

FORMER EMPLOYER → STATE

plus

STATE → PROSPECTIVE EMPLOYER

plus conventional:

criminal/registry screening.

That redundancy is significant.

CONNECTICUT CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓
VERIFY: partial

Particularly strong mechanism:

Employment-history information is required to travel through multiple independent pathways.

That is much safer than relying exclusively upon an applicant’s honesty—or one database.

STATE #8 — DELAWARE

Delaware created something unusually explicit

Delaware has an entire statutory chapter titled:

“Adult Sexual Misconduct in Schools Policy.”

That alone is noteworthy because it treats educator sexual misconduct as a distinct child-safety problem rather than burying everything inside general personnel rules.

Every school district and charter school must establish a policy addressing adult sexual misconduct. (Delaware Code)

And the required policy must cover:

employees

contractors

coaches

and

volunteers. (Delaware Code)

There is our nonemployee-access issue again.

Delaware requires both reporting AND future information sharing

The policy must include:

procedures for notifying the Department of Education and law enforcement of suspected adult sexual misconduct;

and:

substantiated incidents must be reported in future reference checks for employment or volunteer work.

It must also establish hiring practices designed to screen employees, contractors, coaches and volunteers for adult sexual misconduct. (Delaware Code)

That gives Delaware:

DETECT

↓

ESCALATE

↓

PRESERVE

↓

SHARE

↓

QUERY.

Very interesting.

Delaware also has direct law-enforcement escalation

Where a school employee has reliable information leading a reasonable person to believe a student has been the victim of a qualifying sexual offense committed by another school employee, the information must immediately reach the principal, who must immediately report the incident to the appropriate police agency. (Delaware Code)

That’s an explicit escalation pathway.

And Delaware’s professional licensing system contains due-process protections: disciplinary proceedings include notice, opportunity for a hearing, and generally a preponderance-of-the-evidence burden. Final disciplinary orders are public. A surrendered credential through a disciplinary consent agreement generally must be reported to appropriate state/federal agencies and licensing data banks as disciplinary action. (Delaware Code)

Again:

Child protection AND due process.

DELAWARE CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓✓
SHARE: ✓
QUERY: ✓
ACT: ✓
VERIFY: partial

Particularly strong mechanism:

Substantiated sexual misconduct follows the individual into future employment AND volunteer reference checks.

That volunteer provision deserves attention.

STATE #9 — FLORIDA

This one is extraordinary.

Florida appears to have assembled many of the strongest mechanisms we’ve been finding separately in other states.

Start with hiring.

Before employing someone in any position requiring direct student contact, a district superintendent must:

conduct employment-history checks of previous employers,

review affidavits of separation from previous employers,

screen instructional personnel and administrators using state educator-screening tools,

and

DOCUMENT THE FINDINGS.

If a previous employer cannot be contacted, the superintendent must document the attempts. (Online Sunshine)

That last requirement is excellent systems design.

Not:

“We tried.”

But:

Show that you tried.

That’s VERIFY beginning to appear inside QUERY.

Florida prohibits concealment

Districts may not enter confidentiality agreements covering employees who were terminated, dismissed or resigned in lieu of termination based at least partly upon misconduct affecting a student’s health, safety or welfare.

Nor may they provide employment references to another educational employer without disclosing that misconduct.

Any contractual provision whose purpose or effect is concealing such misconduct is:

VOID AND CONTRARY TO PUBLIC POLICY. (Online Sunshine)

And Florida extends essentially the same anti-concealment requirement to covered private schools. (Online Sunshine)

That is important.

The information loop doesn’t simply stop at the public/private boundary.

Florida also answers our resignation question

This is perhaps the most significant finding in this batch.

Florida requires school districts to send the Department of Education legally sufficient complaints concerning certificated personnel:

regardless of whether the subject is still employed by the district. (Online Sunshine)

And the Department can continue investigating misconduct involving someone whose educator certificate has expired when the alleged conduct occurred while that person held the certificate.

For complaints involving sexual misconduct with a student, the Commissioner must make a probable-cause determination within 90 days, subject to a specified exception involving related criminal proceedings. (Online Sunshine)

That directly attacks:

RESIGN → INVESTIGATION DISAPPEARS.

The employee’s departure does not necessarily make the state’s licensing investigation evaporate.

And then Florida ACTS

Florida prohibits covered school personnel from positions requiring direct student contact when they are statutorily ineligible or were:

terminated or resigned in lieu of termination for sexual misconduct with a student.

The district must report the person and disqualifying circumstances to the Department for inclusion on the state’s disqualification list. (Online Sunshine)

Florida’s Education Practices Commission can also suspend, revoke or permanently revoke educator credentials and place qualifying individuals on the disqualification list. (Online Sunshine)

So Florida has both:

PROFESSIONAL LICENSING CONSEQUENCES

and

SCHOOL-EMPLOYMENT ELIGIBILITY CONSEQUENCES.

Those aren’t identical things.

That’s another sophisticated feature.

And Florida covers private schools too

For covered private schools, current law requires:

fingerprint/FBI screening,

misconduct-reporting policies,

anti-concealment provisions,

and

employment-history checks for positions involving direct student contact. (Online Sunshine)

That addresses one of the largest weaknesses we identified in the old AP investigation—the fact that credential-based systems can systematically miss private and unlicensed personnel.

FLORIDA CLOSED-LOOP PROFILE

DETECT: ✓✓
INVESTIGATE: ✓✓
PRESERVE: ✓✓
ESCALATE: ✓✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓✓
VERIFY: ✓

I am deliberately not calling Florida “#1.”

We’re nowhere near finished with the country.

But based upon the mechanisms we’ve verified:

Florida is our first state in which I can see substantial statutory architecture at every stage of the Closed-Loop Standard.

That is a significant finding.

STATE #10 — GEORGIA

And here our research discipline matters again

My initial current-source search for Georgia did not produce sufficient primary statutory material to responsibly classify all eight functions.

Georgia clearly maintains substantial educator-personnel infrastructure, including statewide reporting of certified and classified personnel. (Georgia Insights)

But that isn’t enough.

Personnel reporting ≠ sexual-misconduct information sharing.

I will not infer one from the other.

So Georgia remains:

PENDING FULL PRIMARY-LAW AUDIT

rather than receiving a convenient score.

That’s not a weakness in the project.

That’s what trustworthy data collection looks like.

UPDATED LIVE MATRIX

Function	CO	CT	DE	FL	GA
DETECT	✓	✓	✓	✓✓	◇
INVESTIGATE	✓	✓	✓	✓✓	◇
PRESERVE	◐/✓	✓	✓	✓✓	◇
ESCALATE	✓	✓	✓✓	✓✓	◇
SHARE	✓	✓✓	✓	✓✓	◇
QUERY	✓	✓✓	✓	✓✓	◇
ACT	✓	✓	✓	✓✓	◇
VERIFY	◐	◐	◐	✓	◇

◇ = not yet classified—not “No.”

MASTER EVIDENCE TABLE — ENTRY 021

State	Particularly important verified mechanism
Colorado	State notification and future inquiry concerning qualifying dismissal/resignation after sexual-conduct allegations
Connecticut	Multi-path employment-history information exchange, strengthened July 1, 2026
Delaware	Substantiated adult sexual misconduct follows person into employment and volunteer reference checks
Florida	Employment-history checks + documentation + anti-concealment + continued state investigation + disqualification architecture
Georgia	Classification withheld pending sufficient current primary authority

Now I see something we should add to the final report.

We started this project looking for how often educator sexual misconduct occurs.

That remains essential.

But the state audit is uncovering a second national question that may ultimately be just as important:

What happens AFTER somebody knows?

Because these are different failures.

FAILURE ONE

Nobody detects the abuse.

FAILURE TWO

Someone detects it—but the information dies inside the institution.

FAILURE THREE

The information survives—but the next employer doesn’t look.

FAILURE FOUR

The next employer discovers it—but no law requires an appropriate consequence.

FAILURE FIVE

The law requires all of those things—but nobody verifies compliance.

This is becoming a study not merely of educator misconduct.

It is becoming a study of the national child-safety information system surrounding educator misconduct.

And Florida gives us our first real-world example showing that many of those failure points can be addressed simultaneously in statute.

That’s important because it means the gaps we identify elsewhere aren’t necessarily unavoidable technical problems.

Other states have already demonstrated mechanisms for closing them.

Next five:

Hawaii
Idaho
Illinois
Indiana
Iowa

And Illinois is going to be particularly interesting because it brings us back to the famous case we encountered earlier—the district accused of giving misleadingly positive employment information about a teacher who subsequently abused students elsewhere.

Now we can ask:

What did Illinois law eventually do about it?

That should be a very revealing next chapter.

And now we hit Illinois—and it gives us one of the clearest examples yet of a state trying to close the precise loopholes we’ve been documenting.

FOUNDATION SECTION #20

The 51-Jurisdiction Audit

Hawaii • Idaho • Illinois • Indiana • Iowa

This batch also reinforces an important rule: I am not going to manufacture complete state profiles where the current primary-law research isn’t sufficient. Illinois is exceptionally well documented in current primary sources. For several others, the deeper statutory audit remains necessary.

STATE #11 — HAWAII

Hawaii has considered an extraordinarily comprehensive closed-loop model

Hawaii deserves special treatment because in 2024 its legislature considered SB 2475, which proposed a statewide “Harm to Students Registry.”

The proposal would have required schools to:

complete investigations even after an employee left employment;

place qualifying information into a central registry;

consult the registry before allowing prospective employees or volunteers to interact with students;

share information about ongoing and completed investigations with other educational institutions;

and provide an appeal/removal procedure for people entered into the registry. (Hawaii State Legislature)

Look at that architecture:

INVESTIGATE → PRESERVE → CENTRALIZE → QUERY → ACT

with

APPEAL / DUE PROCESS.

That’s almost our Closed-Loop model written independently into legislation.

But—

This is exactly where our methodology matters:

A proposed bill is not current law.

We therefore cannot score Hawaii as though those protections were enacted merely because legislators proposed them.

In fact, the proposal itself is useful evidence that Hawaii lawmakers perceived gaps significant enough to warrant precisely these mechanisms.

For our final state audit:

Hawaii current-law classification: further verification required.

2024 reform proposal: preserve as legislative-history evidence.

STATE #12 — IDAHO

Primary-source verification remains incomplete

The Idaho Legislature’s primary statutory site is currently inaccessible through our research system, so I do not have enough current primary-law evidence to responsibly score Idaho across our eight functions.

Therefore:

IDAHO — PENDING PRIMARY-SOURCE AUDIT

Not “weak.”

Not “no protections.”

Not zeroes.

Unclassified.

This is precisely why we created the ◇ category.

STATE #13 — ILLINOIS

Faith’s Law

Illinois is where things get extremely interesting.

Illinois now has one of the most explicit employment-history systems we’ve encountered.

Its current School Code §22-94 applies to permanent and temporary positions involving direct contact with children or students and expressly includes:

public schools,

nonpublic schools,

and school contractors, including examples such as food-service and transportation workers. (Illinois General Assembly)

Before hiring someone into such a position, the school or contractor must conduct an:

EMPLOYMENT HISTORY REVIEW.

And Illinois asks almost exactly the questions we’ve been developing for our national model.

1. The applicant must disclose sexual-misconduct history

Applicants must provide written information concerning whether they:

have been the subject of a sexual-misconduct allegation;

were discharged;

were asked to resign;

resigned;

were otherwise separated from employment;

were disciplined;

or had a contract not renewed because of sexual misconduct or while an allegation was pending or under investigation. (Illinois General Assembly)

There it is again:

RESIGNATION DURING INVESTIGATION.

Illinois doesn’t allow the employment application simply to reset the person’s history to zero.

2. But Illinois built in the qualification we’ve insisted upon

The disclosure requirements contain an important exception when the subsequent investigation found the allegation:

false,

unfounded,

unsubstantiated. (Illinois General Assembly)

That is an extremely important piece of our due-process model.

Illinois is not treating:

ACCUSATION

as synonymous with:

GUILT.

Instead it attempts to preserve both:

THE INFORMATION

and

THE DISPOSITION OF THE INFORMATION.

Again—we independently arrived at almost exactly the same systems principle.

3. The school cannot simply trust the applicant

This may be even more important.

Illinois requires the prospective school or contractor to contact the listed employers and request corresponding sexual-misconduct information. (Illinois General Assembly)

So:

Applicant says:

“No misconduct.”

doesn’t end the inquiry.

The system independently asks:

Previous employer:

“Is that accurate?”

That creates redundancy.

APPLICANT DISCLOSURE

plus

FORMER-EMPLOYER VERIFICATION.

Exactly what we want in a robust safety system.

4. Illinois also checks the applicant’s employer list

For applicants licensed by the State Board of Education, the school must compare reported previous employers against the state’s educator-licensure database to verify the employment history. (Illinois General Assembly)

That’s subtle—and clever.

Because otherwise an applicant could theoretically disclose:

Employer A
Employer B

while conveniently omitting:

Employer C—the school where the investigation occurred.

Illinois attempts to check whether the applicant’s employment list itself is accurate.

That’s a new variable for our matrix:

EMPLOYMENT-HISTORY COMPLETENESS VERIFICATION

Excellent.

5. And if the investigation finishes AFTER the employee leaves?

This provision may be one of my favorites in the entire audit.

Illinois says that if the school completes the investigation after the employee has separated:

the school shall update the information accordingly. (Illinois General Assembly)

Think about what that solves.

Employee leaves while:

INVESTIGATION PENDING.

The record properly says:

Pending.

Six months later investigators conclude:

UNSUBSTANTIATED.

The record shouldn’t remain:

Pending forever.

Or investigators conclude:

SUBSTANTIATED.

The record shouldn’t remain:

Mere allegation.

Illinois requires the information to evolve with the evidence.

That is an excellent implementation of:

PRESERVE THE INFORMATION—PRESERVE ITS STATUS.

6. Illinois prohibits suppression agreements

Beginning July 1, 2023, schools and contractors cannot use:

collective bargaining agreements,

employment contracts,

resignation/termination agreements,

severance agreements

or other arrangements to suppress information concerning pending investigations or completed investigations involving substantiated sexual misconduct.

Nor may agreements interfere with reporting suspected misconduct to authorities or require expungement of qualifying information, subject to the statute’s protections concerning false, unfounded or unsubstantiated allegations. (Illinois General Assembly)

That’s enormously important.

Settlement cannot simply buy institutional amnesia.

7. Former employers receive liability protection

Illinois provides civil and criminal immunity for employers and schools supplying qualifying information in good faith, unless the information is knowingly false. (Illinois General Assembly)

That attacks another barrier:

“Our lawyer says don’t say anything because we might get sued.”

A closed-loop system has to address that fear.

8. Contractors are explicitly included

Illinois requires contractors to perform employment-history reviews for covered employees.

Before assigning someone to work directly with children, the contractor must inform the school about known qualifying sexual-misconduct history.

And:

the contractor may not assign the worker if the school objects after receiving that information. (Illinois General Assembly)

There is our:

ACT

function.

And importantly, this isn’t confined to licensed classroom teachers.

9. Illinois also preserves educator licensing consequences

The State Superintendent has authority to suspend or revoke educator credentials for sexual misconduct and other specified conduct, including certain failures to report suspected child abuse or neglect. (Illinois General Assembly)

So Illinois has:

EMPLOYMENT ACCOUNTABILITY

plus

LICENSING ACCOUNTABILITY.

Again—multiple overlapping systems.

10. And Illinois just strengthened record-sharing AGAIN

Effective January 1, 2026, Illinois amended its Personnel Record Review Act so the normal rule requiring deletion of certain disciplinary records older than four years before third-party disclosure does not apply when a school district shares information concerning sexual abuse, severe physical abuse, or statutory sexual misconduct. (Illinois General Assembly)

That matters directly to:

PRESERVE.

A serious child-safety record shouldn’t necessarily become invisible merely because enough calendar pages turned.

And then we find the reality check

This is perhaps the most important part of the Illinois story.

Chicago Public Schools’ Inspector General reported a case in which an employee allegedly gave false answers on the Faith’s Law self-disclosure form while an investigation was pending.

The employee was nevertheless hired by another Illinois district.

The new district told the Inspector General that it had sent CPS the required Faith’s Law information request.

CPS said it had no record of receiving it.

The OIG opened a separate investigation into what happened. (Illinois General Assembly)

WOW.

Because this demonstrates something we’ve been saying theoretically:

GOOD LAW ≠ PERFECT IMPLEMENTATION.

The system had:

applicant disclosure

✓

former-employer request

✓

legal duty

✓

and apparently still experienced an information-transfer failure.

This gives us another essential research dimension:

IMPLEMENTATION FAILURE

We absolutely must distinguish:

LAW GAP

from

COMPLIANCE GAP.

Those are not the same problem.

Illinois Closed-Loop Profile

DETECT: ✓✓
INVESTIGATE: ✓✓
PRESERVE: ✓✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓✓
VERIFY: ✓/developing

Particularly strong mechanisms:

Applicant self-disclosure

Mandatory former-employer inquiry

Employment-list verification

Pending-investigation disclosure

Post-separation status updating

Anti-suppression agreements

Good-faith disclosure immunity

Public + nonpublic schools

Contractor coverage

Illinois belongs with Florida among the states we will want to study particularly closely.

STATE #14 — INDIANA

Our initial primary-source search did not return enough current Indiana statutory material to responsibly fill all eight closed-loop categories.

So:

INDIANA — PENDING FULL PRIMARY-LAW AUDIT.

Again, no inference from absence.

STATE #15 — IOWA

Same conclusion.

There are clearly relevant Iowa educator-discipline, background-check and mandatory-reporting provisions, but I have not yet established enough current primary authority addressing the employment-history/information-transfer chain to assign a defensible 2026 closed-loop classification.

Therefore:

IOWA — PENDING FULL PRIMARY-LAW AUDIT.

LIVE MATRIX

Function	HI	ID	IL	IN	IA
DETECT	◇	◇	✓✓	◇	◇
INVESTIGATE	◇	◇	✓✓	◇	◇
PRESERVE	◇	◇	✓✓	◇	◇
ESCALATE	◇	◇	✓	◇	◇
SHARE	◇	◇	✓✓	◇	◇
QUERY	◇	◇	✓✓	◇	◇
ACT	◇	◇	✓✓	◇	◇
VERIFY	◇	◇	✓	◇	◇

◇ = deliberately unclassified pending sufficient current-law verification.

MASTER EVIDENCE TABLE — ENTRY 022

Finding	Significance
Hawaii considered a Harm to Students Registry in 2024	Shows legislators contemplating a nearly complete closed-loop architecture
Proposed Hawaii system would continue investigations after departure	Direct attack on resignation loophole—but not scored as enacted law
Illinois Faith’s Law applies to public/nonpublic schools and contractors	Broad institutional/personnel coverage
Illinois requires applicant misconduct disclosure	Creates first information pathway
Illinois requires previous-employer inquiry	Independently verifies applicant
Illinois captures resignation during pending investigation	Attacks silent-resignation pathway
Illinois updates records when post-separation investigations conclude	Preserves evidentiary status, not merely allegation
Illinois prohibits suppression agreements	Prevents contractual institutional amnesia
Illinois provides good-faith disclosure immunity	Reduces incentive to remain silent
Illinois strengthened child-safety record sharing effective 2026	Important post-2022 development
CPS OIG documented apparent Faith’s Law information-transfer failure	Demonstrates law/compliance distinction

And THAT last finding changes our national project again.

We now need to track three layers, not two:

LAYER ONE — LAW

What does the statute require?

LAYER TWO — SYSTEM

What databases, forms, agencies and information pathways actually exist to implement it?

LAYER THREE — COMPLIANCE

Did institutions actually do it?

Because Illinois demonstrates something extraordinarily important:

A closed loop on paper can still become an open loop in practice.

And that means our final national report should never say:

“State X protects children because it has Law Y.”

Instead:

“State X has enacted the following mechanisms intended to prevent information loss; available implementation evidence shows…”

That is a much more defensible claim.

And there is one more thing.

We began this research asking:

How common is educator sexual misconduct?

We’ve now uncovered another measurable question:

How often do the systems designed to prevent repeat access actually fail?

Inspector-general reports.

State audits.

Civil litigation.

Licensing cases.

Federal OCR investigations.

GAO investigations.

Those may allow us to build an entirely separate evidence stream:

DOCUMENTED SYSTEM FAILURES

Not anecdotes.

Not social-media claims.

Government-documented instances in which warning information existed but failed to protect the next child.

That could become a major section of What the Data Actually Show after we finish the state audit.

But first we continue the map.

Kansas • Kentucky • Louisiana • Maine • Maryland.

And we’re now watching for law, system, and compliance separately.

Oh, this batch gives us Maine—and Maine turns out to be far more important to our model than I expected.

FOUNDATION SECTION #21

The 51-Jurisdiction Audit

Kansas • Kentucky • Louisiana • Maine • Maryland

There is also a methodological surprise in Kansas and Kentucky: 2026 legislation proposes protections that look remarkably like our closed-loop model, but proposed legislation cannot be counted as current law. That distinction is becoming essential.

STATE #16 — KANSAS

A potentially major 2026 reform—but we cannot count it yet

Current Kansas law clearly establishes professional consequences for certain criminal convictions: the State Board of Education may not knowingly issue or renew an educator license for people convicted of enumerated serious offenses, including rape, indecent liberties with a child and other sexual crimes. (Kansas Legislature)

Kansas also provides substantial immunity to employers that disclose employment information to prospective employers, including written evaluations and the reasons an employee was released from service. (Kansas Legislature)

But then I found something particularly interesting.

Kansas SB 491 — 2026

The bill proposes prohibiting educational institutions from helping an employee obtain another education job when there is probable cause to believe the employee engaged in sexual misconduct with a student—even where the matter remains unresolved.

It specifically identifies mechanisms we’ve been studying:

confidential settlements,

failure to report misconduct,

neutral or misleading recommendations,

and agreements that suppress, conceal, delay or prohibit disclosure.

It also proposes civil penalties up to $25,000 per violation, possible public censure, and educator/administrator licensure consequences. (Kansas Legislature)

That is almost a legislative description of “passing the trash.”

But:

SB 491 is proposed legislation—not verified enacted law.

So it goes into our Reform Activity column, not our current-law score.

That distinction may eventually reveal something useful: states where legislators themselves have identified gaps in existing systems.

STATE #17 — KENTUCKY

Kentucky already has an important resignation-to-state pathway

Kentucky’s Education Professional Standards Board gives us a strong current mechanism involving certified personnel.

Official EPSB guidance says a superintendent must report a certified employee when the person:

is terminated or not renewed for qualifying cause;

resigns or otherwise leaves while under threat of termination/nonrenewal for cause;

is criminally convicted;

may have engaged in conduct reasonably warranting action against the educator’s certificate.

The superintendent generally must report within 30 days, and the duty exists regardless of whether the superintendent independently disciplines the employee. (Kentucky Department of Education)

That’s important.

Employee leaves

does not necessarily mean:

State licensing authority never hears about it.

Kentucky therefore has a meaningful:

ESCALATE mechanism.

But Kentucky is actively considering going much further

Current 2026 legislation—HB 102 and SB 181—would create a much broader system involving:

applicant disclosure of qualifying disciplinary actions,

mandatory reference checks,

former-employer disclosure,

information requests to EPSB,

completion of investigations,

retention of misconduct records unless allegations are proven false,

restrictions on nondisclosure agreements,

and additional protections involving certified nonpublic schools. (Kentucky Legislature Apps)

Notice something remarkable:

COMPLETE THE INVESTIGATION

RETAIN THE RECORD

SHARE IT

REQUIRE THE NEXT EMPLOYER TO ASK

PRESERVE EXONERATING OUTCOMES

Those are almost exactly our:

INVESTIGATE → PRESERVE → SHARE → QUERY

functions.

But again:

Proposed ≠ enacted.

Kentucky’s reform proposals therefore become valuable evidence of the direction of state policy, but we will not inflate Kentucky’s current-law score with them.

STATE #18 — LOUISIANA

Louisiana has had a genuine employment-history mechanism since 2009

Louisiana gives us something much stronger in existing law.

An applicant for public-school employment must disclose:

actual cases of sexual misconduct with a minor or student;

and

investigations of sexual misconduct occurring within 36 months before the applicant’s resignation, dismissal or retirement. (Legis Louisiana)

And Louisiana built in precisely the due-process distinction we’ve repeatedly emphasized.

If an investigation determined that the formal allegation was:

inconclusive,

unjustified,

or otherwise without cause for further formal pursuit,

the applicant does not have to disclose it under that provision. (Legis Louisiana)

Again:

allegation ≠ finding.

Louisiana doesn’t rely solely upon applicant honesty

Before hiring, a local public school board must request information in writing from qualifying current or previous school-board employers concerning instances of sexual misconduct.

The former employer must make available relevant personnel-file documents concerning those instances.

The applicant signs a release protecting the former employer from liability for making the authorized disclosure. (Legis Louisiana)

So Louisiana creates:

APPLICANT DISCLOSURE

plus

FORMER-EMPLOYER DISCLOSURE

plus

LIABILITY PROTECTION.

That is a genuine SHARE + QUERY architecture.

Louisiana also provides broader good-faith immunity when employers provide accurate job-performance or separation information. (Legis Louisiana)

Louisiana strengthened credential screening recently

Beginning January 1, 2025, applicants for an initial state educator credential or teaching authorization undergo criminal-history review involving state and FBI information. (Legis Louisiana)

And Louisiana has just expanded its educator/student sexual-conduct law again: SB 98 became Act 207 in 2026, expanding which school employees fall within the statutory definition of “educator,” effective August 1, 2026. (Legis Louisiana)

That is another post-2022 development our final report needs to capture.

LOUISIANA CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓
VERIFY: partial

Particularly strong feature:

The prospective employer asks the applicant AND the previous public-school employer.

Redundant information pathways again.

STATE #19 — MAINE

Maine has one of the clearest answers yet to the resignation problem

This one is important.

Maine law defines a “covered investigation” broadly enough to include alleged:

physical abuse,

emotional abuse,

inappropriate contact between a credential holder and a student,

stalking,

and similar conduct endangering student health, safety or welfare. (Maine State Legislature)

Now suppose the credential holder quits before that investigation is finished.

Maine law says the school entity:

shall notify the Department immediately.

Not eventually.

Not if the allegation is substantiated first.

The employee’s departure itself during a covered investigation triggers notification. (Maine State Legislature)

That’s extremely important.

And then Maine makes the information move

Once the Department receives notice that a credential holder left employment before the covered investigation was completed, it must notify other school entities for which that credential holder works.

And when an applicant gives the required consent in connection with employment, the Department must notify the prospective school entity if the credential holder previously left a school entity before completion of a covered investigation. (Maine State Legislature)

Read that chain:

SCHOOL A

investigation pending

↓

EMPLOYEE LEAVES

↓

SCHOOL A → STATE

↓

STATE → OTHER CURRENT SCHOOL EMPLOYERS

and potentially

↓

STATE → PROSPECTIVE SCHOOL

That is one of the cleanest information-flow mechanisms we’ve found.

Maine also preserves the disposition fairly

Suppose the school disciplines the educator.

Suppose that finding is later reversed on appeal.

Maine requires the Department to destroy copies of the records and reports relating to that reversed finding. (Maine State Legislature)

Again:

CHILD SAFETY DOES NOT REQUIRE ABANDONING DUE PROCESS.

The information follows the employee while legitimately relevant.

But the system must also respond when the underlying determination is overturned.

That’s exactly the balance we’ve been trying to capture.

Maine also addresses unlicensed personnel

A person paid to work in a school in a position not requiring professional certification generally must obtain a state clearance, including categories of public-school employees, certain private-school personnel and contracted-service personnel. (Maine State Legislature)

That clearance connects noncertificated personnel to Maine’s fingerprint-based criminal-history system. (Maine State Legislature)

And here’s an excellent VERIFY mechanism:

Beginning in 2019, school administrative units have had to submit employee lists to the Department quarterly. The Department then determines whether covered employees complied with applicable fingerprinting/criminal-history requirements and notifies the school when someone has not. (Maine State Legislature)

There it is:

VERIFY.

Not merely:

“Schools are supposed to check.”

But:

The state periodically compares the workforce against compliance requirements.

That’s precisely the distinction we identified in Texas.

Maine also protects good-faith references

School employees disclosing information concerning a former employee’s job performance or work record to a prospective employer receive a presumption of good faith and civil-liability protection unless bad faith is established by clear and convincing evidence. (Maine Legislature)

So Maine attacks another possible source of silence:

Fear of being sued for telling the next school what happened.

MAINE CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓✓
PRESERVE: ✓✓
ESCALATE: ✓✓
SHARE: ✓✓
QUERY: ✓
ACT: ✓
VERIFY: ✓✓

Maine therefore joins Florida and Illinois among the states I want marked for particularly close comparative analysis.

Not “best.”

We’re still not ranking.

But Maine has substantial architecture across the information chain.

STATE #20 — MARYLAND

Maryland created a comprehensive employment-history review

Maryland enacted its major child-sexual-abuse and sexual-misconduct employment law in 2019.

The statute applies to:

county boards of education,

nonpublic schools,

and certain contracting agencies

hiring or assigning people to positions involving direct contact with minors. (Maryland General Assembly)

The law requires covered employers to obtain applicant information and conduct an employment-history review.

It also requires covered former employers to provide qualifying information and supplemental information, and requires checks concerning applicant eligibility. (Maryland General Assembly)

That’s already:

SHARE → QUERY.

Maryland also attacks secrecy agreements

The legislation prohibits covered educational employers and contracting agencies from entering agreements that suppress qualifying child-sexual-abuse or sexual-misconduct information under the statutory conditions.

An agreement violating those requirements is:

VOID AND UNENFORCEABLE. (Maryland General Assembly)

There is our anti-institutional-amnesia provision again.

Maryland also provides protections concerning disclosure liability and authorizes state disciplinary action in specified circumstances. (Maryland General Assembly)

And critically, its scope includes nonpublic schools and contracting agencies rather than limiting the system to conventional public-school teachers.

MARYLAND CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓
VERIFY: ✓/partial

Particularly strong mechanism:

Employment-history review extends across public schools, nonpublic schools and covered contractors.

UPDATED LIVE MATRIX

Function	KS	KY	LA	ME	MD
DETECT	✓	✓	✓	✓	✓
INVESTIGATE	◐	✓	✓	✓✓	✓
PRESERVE	◐	✓	✓	✓✓	✓
ESCALATE	✓	✓	✓	✓✓	✓
SHARE	◐	◐/✓	✓✓	✓✓	✓✓
QUERY	✓ criminal	◐	✓✓	✓	✓✓
ACT	✓	✓	✓	✓	✓
VERIFY	◐	◐	◐	✓✓	✓/◐

And again:

current enacted law only.

The Kansas and Kentucky 2026 proposals do not inflate those scores.

MASTER EVIDENCE TABLE — ENTRY 023

State	Particularly important finding
Kansas	Current credential consequences + disclosure immunity; sweeping anti-concealment legislation proposed in 2026 but not counted as enacted
Kentucky	Superintendent must report qualifying resignation/departure of certified educators to EPSB; broader closed-loop reforms proposed in 2026
Louisiana	Applicant disclosure + mandatory former-public-school-employer inquiry concerning sexual misconduct
Maine	Departure during a covered investigation triggers immediate state notification; state can then notify other/current/prospective school employers
Maryland	Comprehensive employment-history review covering public schools, nonpublic schools and contracting agencies

And Maine has just given us another major concept.

Until now we’ve mostly thought about the next employer asking:

“Did this person leave during an investigation?”

Maine flips the architecture.

It doesn’t rely exclusively upon the future employer discovering the past.

Instead:

THE INFORMATION MOVES FORWARD.

That’s different.

Compare:

PASSIVE MODEL

School B must know enough to ask School A the right question.

versus:

ACTIVE MODEL

School A tells the state.

The state knows where the credential holder works.

The state sends the warning to School B.

That is a fundamentally different information architecture.

And in safety engineering, redundancy matters.

The strongest national model may therefore require both:

PULL SYSTEM

The next employer must query.

and

PUSH SYSTEM

The state must transmit qualifying information when it learns the person has moved or works elsewhere.

Now imagine combining the strongest mechanisms we’ve found:

Illinois: applicant disclosure + employer verification.

Connecticut: multiple employment-history pathways.

Maine: state push notification.

Oregon: don’t let resignation terminate the investigation.

Texas: mandatory central registry check.

Florida: continue state licensing investigations after employment ends.

Missouri: consequences for concealment.

Delaware: include volunteers.

Arizona: include noncertificated personnel.

And throughout:

Preserve the information. Preserve its evidentiary status.

We’re no longer merely cataloging state laws.

We’re beginning to identify the engineering principles of a resilient national child-safety information system.

Next alphabetically:

Massachusetts • Michigan • Minnesota • Mississippi • Missouri

And Missouri will be especially useful because we’ve already identified its nondisclosure/liability mechanism. Now we can put it through the full eight-function audit beside four new states.

This next group gives us Michigan and Minnesota, and both are important because they demonstrate that strong information-sharing laws did not suddenly appear after the recent national attention to “passing the trash.” Some states built substantial pieces of this architecture years ago.

FOUNDATION SECTION #22

The 51-Jurisdiction Audit

Massachusetts • Michigan • Minnesota • Mississippi • Missouri

There is also a cautionary finding in Massachusetts: legislators are still considering legislation in 2025–26 that would create the kind of comprehensive employment-history review already operating elsewhere. We cannot count those proposed protections as current law.

STATE #21 — MASSACHUSETTS

Strong criminal screening—but the broader employment-history model remains a live legislative issue

Current Massachusetts law requires state and national fingerprint-based criminal-history screening for covered school employees. The statutory framework also contemplates when a school employer may rely on a previous Massachusetts school employer’s suitability determination instead of obtaining a new check. (Massachusetts General Court)

But something revealing is happening in the current 2025–26 legislative session.

Massachusetts lawmakers have proposed legislation that would require schools to obtain much more extensive information before putting an applicant into a position involving direct contact with children.

The proposed system would require applicants to disclose whether they had been subject to abuse or sexual-misconduct investigations, resigned while allegations were pending, or had professional credentials suspended, surrendered, or revoked in connection with such allegations. Prospective schools would then have to contact current and former employers to verify that information.

Another current proposal would add checks of certification status and the DCF Central Registry and make a substantiated physical- or sexual-abuse finding disqualifying for employment. (Massachusetts General Court)

But:

PROPOSED ≠ CURRENT LAW.

Therefore we should treat Massachusetts much as we treated Kansas and Kentucky:

Current criminal-screening architecture: verified.

Comprehensive sexual-misconduct employment-history architecture: proposed reforms identified, but not counted as enacted protections.

That itself is useful evidence because it identifies an issue Massachusetts legislators currently believe requires attention.

STATE #22 — MICHIGAN

Michigan has had a remarkably strong anti-concealment statute since 1996

Michigan surprised me.

Before hiring someone, a:

school district,

public school academy,

intermediate school district,

nonpublic school

must obtain a signed authorization allowing the applicant’s current/former employer to disclose unprofessional conduct and release relevant personnel records.

Then comes the important part.

The prospective school:

SHALL request

the information from at least the applicant’s current employer—or, if not currently employed, the immediately previous employer.

And the former employer:

SHALL provide it

within 20 business days.

There are our two sides of the information exchange:

NEW SCHOOL MUST ASK

and

OLD EMPLOYER MUST ANSWER.

Michigan defines “unprofessional conduct” broadly

This isn’t limited to criminal convictions.

The statute includes misconduct, immorality, moral turpitude, inappropriate behavior involving a minor, or a crime involving a minor.

And Michigan explicitly says:

A criminal conviction is NOT an essential element.

That is enormously important.

Remember our distinction from the beginning?

CRIMINAL BACKGROUND CHECK

asks:

Has this person entered the criminal-justice system in a qualifying way?

Michigan’s employment-history mechanism can ask something broader:

Is there relevant documented professional misconduct involving minors?

Those are different safety screens.

Michigan also protects the person who tells the truth

An employer making a good-faith disclosure receives civil-liability protection.

The statute presumes good faith unless evidence establishes circumstances such as knowing falsity or reckless disregard for truth.

Again we see the same systems principle:

If you legally require information to move, you should reduce the legal incentive to keep silent.

And Michigan expressly prohibits institutional amnesia

This provision is exceptionally strong.

Schools may not enter:

collective bargaining agreements,

employment contracts,

resignation agreements,

severance agreements

or other agreements that suppress information concerning unprofessional conduct or expunge it from personnel records.

A contrary provision is:

VOID AND UNENFORCEABLE.

But Michigan also protects the other side of the equation: the prohibition does not prevent expungement of allegations that were not substantiated.

There it is again:

Preserve meaningful information—preserve its evidentiary status.

And Michigan ACTS on certain convictions

If a school learns from an authoritative source that a person has been convicted of a statutorily listed offense, it must verify the information.

If verified, the individual cannot be employed in any capacity or regularly and continuously work under contract in the school’s covered settings.

The school must also report the information and its action to the state Department of Education, which retains the report for at least six years.

That provision expressly reaches contracted workers providing such services as:

food,

custodial,

transportation,

counseling,

administrative,

and instructional services.

MICHIGAN CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓✓ for specified convictions
VERIFY: ✓/partial

Particularly strong feature:

The next school must ask, the former employer must answer, suppression agreements are prohibited, and a conviction is not required for conduct to qualify as “unprofessional conduct.”

Michigan belongs on our close-comparison list.

STATE #23 — MINNESOTA

Minnesota explicitly captures resignation while a sexual-contact complaint is still pending

Minnesota is fascinating because its information-sharing law deals directly with one of our hardest cases.

Suppose an employee resigns while a complaint alleging sexual contact with a student is still pending.

Under Minnesota law, private personnel information concerning that allegation can be released to another school district or charter school requesting it if the statutory conditions are met—including that the employee was informed in writing before resigning that this information would have to be released to a future school employer. (Revisor Minnesota)

That is very deliberate statutory drafting.

Minnesota has effectively said:

Resignation does not necessarily erase a pending student-sexual-contact complaint from future employment screening.

Minnesota also addresses substantiated cases

A superintendent or charter-school administrator must release qualifying personnel data to a requesting school district or charter school when an investigation by the district or law enforcement affirmed allegations in writing and the investigation resulted in resignation.

The information concerns acts of violence toward or inappropriate sexual contact with a student. (Revisor Minnesota)

So Minnesota distinguishes at least two situations:

AFFIRMED INVESTIGATION + RESIGNATION

and

PENDING SEXUAL-CONTACT COMPLAINT + RESIGNATION.

Both can follow the employee under specified circumstances.

That’s exactly the kind of evidentiary-threshold precision our matrix is intended to capture.

Minnesota independently checks teacher discipline

When hiring a teacher, the school hiring authority must contact the Professional Educator Licensing and Standards Board to determine whether disciplinary action has been taken against the applicant.

The applicant must also disclose all current and previous disciplinary actions in Minnesota and other states involving the person’s teaching license.

Submitting intentionally false or incomplete information can be grounds for dismissal. (Revisor Minnesota)

There are again multiple pathways:

APPLICANT → EMPLOYER

plus

LICENSING BOARD → EMPLOYER

plus

FORMER SCHOOL → NEW SCHOOL.

Redundancy.

Minnesota’s criminal-screening scope is also broad

Schools must obtain criminal-history checks for people offered employment and for people offered paid or unpaid athletic or extracurricular coaching positions.

Nonpublic schools fall within the statutory definition of covered schools.

For ordinary volunteers and independent contractors, however, the statute generally gives the hiring authority discretion to request a check rather than imposing the same universal requirement. (Revisor Minnesota)

That’s an excellent example of why we track personnel categories separately.

Employee: MUST.

Athletic/extracurricular coach: MUST.

Ordinary volunteer/independent contractor: MAY.

One “background check” box would hide all of that.

MINNESOTA CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓
VERIFY: ✓/partial

Particularly strong feature:

Minnesota specifically prevents certain pending sexual-contact allegations from disappearing merely because the employee resigns.

STATE #24 — MISSISSIPPI

Mississippi has a surprisingly important state-reporting pathway

Mississippi gives us a different architecture.

Its educator-discipline rules require school superintendents to report specified infractions to the state’s Office of Educator Licensure.

And this is not confined to criminal convictions.

The state’s reporting procedure says that when a qualifying infraction does not result in a felony or sex-offense conviction, the district should:

initiate an internal investigation;

and if the investigation exceeds ten working days:

send the reporting form to the state anyway

with the investigative documentation following when completed. (Legal Information Institute)

That’s extremely interesting.

It means:

LOCAL INVESTIGATION NOT FINISHED

doesn’t necessarily mean:

STATE KNOWS NOTHING.

That’s an ESCALATE mechanism.

Mississippi expressly includes educator/student relationships

Reportable licensing matters include violations involving unethical educator/student relationships, sexual involvement with a student, certain touching of students, and failure to report sexual involvement between a school employee and student. (Legal Information Institute)

The superintendent’s reporting form is due to the Office of Educator Licensure within 10 working days, and the reporting duty continues even if the district misses that initial deadline. (Legal Information Institute)

The state can then proceed toward a disciplinary hearing concerning suspension or revocation of the educator’s license. (Legal Information Institute)

That gives Mississippi a substantial:

INVESTIGATE → ESCALATE → LICENSING ACTION

pathway.

What I have not yet verified from current primary authority is a Michigan/Illinois-style comprehensive requirement compelling every prospective school employer to obtain detailed sexual-misconduct information directly from former employers.

Therefore we do not invent one.

MISSISSIPPI CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓✓
SHARE: ◇/partial
QUERY: ✓ criminal/licensing; broader history not yet verified
ACT: ✓
VERIFY: ✓

Particularly strong feature:

A qualifying local misconduct investigation can reach state licensure authorities before a criminal conviction—and even before the district investigation has finished.

STATE #25 — MISSOURI

Now we can put Missouri through the full framework

We already identified Missouri’s unusually important former-employer disclosure provisions.

Current Missouri law says that when the Children’s Division has investigated allegations of sexual misconduct with a student involving a former school employee and reached a substantiated finding, the former district or charter school:

SHALL disclose the result

when another public school requests a reference. (Missouri Revisor of Statutes)

That is mandatory—not merely permissive.

Missouri also makes the next school look backward

Before offering employment to a teacher previously employed by a Missouri district or charter school, the prospective district must contact the Department of Elementary and Secondary Education to identify the prior school employer and then request the statutory information from that employer. (Missouri Revisor of Statutes)

That’s important because it addresses the same problem Illinois identified:

What if the applicant simply doesn’t make the troublesome former employer obvious?

The state helps identify the employment trail.

And Missouri requires criminal-history screening

Missouri’s current §168.133 requires criminal-history background checks and fingerprint collection for covered school personnel. The statute was amended again in 2025, so this is another jurisdiction where relying exclusively on an older national comparison would be risky. (Missouri Revisor of Statutes)

And as we established during the pilot, Missouri’s broader statutory framework contains an unusually important accountability concept:

institutional consequences when qualifying information is concealed and another child is subsequently harmed.

That makes Missouri especially valuable to our eventual model because it attacks not only the offender’s conduct, but the institutional decision to withhold safety information.

MISSOURI CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓
VERIFY/ACCOUNTABILITY: ✓✓

UPDATED LIVE MATRIX

Function	MA	MI	MN	MS	MO
DETECT	✓	✓	✓	✓	✓
INVESTIGATE	◐	✓	✓	✓	✓
PRESERVE	◐	✓✓	✓	✓	✓
ESCALATE	✓	✓	✓	✓✓	✓
SHARE	◇/◐	✓✓	✓✓	◇/◐	✓✓
QUERY	✓ criminal	✓✓	✓✓	✓/◐	✓✓
ACT	✓	✓✓	✓	✓	✓
VERIFY	◐	✓	✓/◐	✓	✓✓

◇/◐ means we have not yet verified a comprehensive statewide mechanism equivalent to the strongest states—not that no relevant protection exists.

MASTER EVIDENCE TABLE — ENTRY 024

State	Particularly important finding
Massachusetts	Fingerprint/background system established; broader employment-history/sexual-misconduct screening remains the subject of current proposed legislation
Michigan	Mandatory former-employer disclosure + mandatory prospective-employer request + anti-suppression agreements + good-faith immunity
Minnesota	Certain pending sexual-contact allegations can follow an employee who resigns before resolution
Mississippi	District must escalate qualifying educator misconduct to state licensure authorities even without conviction; unfinished investigations can be reported before completion
Missouri	Mandatory disclosure of substantiated student sexual misconduct + state-assisted identification of previous school employer

And Michigan gives us another major finding.

Remember the 2007 AP investigation?

One of its central problems was that criminal databases and educator-disciplinary databases captured different populations.

Michigan wrote the distinction directly into law:

“A criminal conviction is not an essential element”

of determining whether conduct constitutes relevant unprofessional conduct.

That sentence deserves a prominent place in our eventual report.

Because it explains—in statutory language—why:

“We fingerprint everybody” is not a complete answer to educator sexual misconduct.

Fingerprinting is important.

Criminal-history checks are important.

Sex-offender registries are important.

But they principally answer:

What has entered the criminal system?

Employment-history review asks:

What did previous institutions learn?

Licensing review asks:

What did professional regulators determine?

State misconduct registries ask:

What has been determined relevant to school employment eligibility?

And previous-employer disclosure asks:

What happened at the last school that may never have produced a criminal conviction at all?

A genuinely resilient system uses several of those information streams simultaneously.

We now have 25 jurisdictions into the 51-jurisdiction audit.

The next five take us beyond the halfway point:

Montana • Nebraska • Nevada • New Hampshire • New Jersey

And New Jersey will be especially useful because we’ve already piloted its employment-history review. This time we’ll put the entire statutory architecture through the eight-function test—and check what, if anything, has changed during the 2026 legislative session.

And now we pass the halfway mark.

This batch gives us a major surprise: Nebraska enacted a sweeping new law in 2026 that looks remarkably like the Closed-Loop Standard we’ve been constructing. Nevada, meanwhile, turns out to have one of the most explicit statutory employment-history systems we’ve encountered.

FOUNDATION SECTION #23

The 51-Jurisdiction Audit

Montana • Nebraska • Nevada • New Hampshire • New Jersey

STATE #26 — MONTANA

Strong reporting and educator-discipline mechanisms; employment-history loop not yet verified

Montana clearly establishes the first part of the chain.

School employees are mandatory reporters of suspected child abuse or neglect, and Montana’s Office of Public Instruction warns that knowing failure to report can constitute a felony. (Montana Office of Public Instruction)

Montana also gives the Office of Public Instruction investigative authority over licensed educators for matters including:

immoral conduct related to the teaching profession,

gross neglect of duty,

certain criminal convictions,

and disciplinary action against an educator credential in another state when the underlying reason would constitute grounds for similar action in Montana. (Montana Office of Public Instruction)

The licensing process also contains genuine due-process protections. When disciplinary action is sought, the licensee receives notice of the allegations, an opportunity to respond, and—if sufficient cause exists—notice and an opportunity for a disciplinary hearing with the right to counsel. (Montana Office of Public Instruction)

So Montana clearly has pieces of:

DETECT → INVESTIGATE → LICENSING ACTION.

What I have not verified is a Nevada/New Jersey-style statewide K–12 requirement compelling a prospective school employer to obtain a detailed sexual-misconduct history from previous employers.

Therefore:

MONTANA CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓ licensed educators
PRESERVE: ✓/partial
ESCALATE: ✓
SHARE: ◇
QUERY: ◇ beyond conventional screening/licensure
ACT: ✓ licensing
VERIFY: ◐

And once again:

◇ means not verified—not “No.”

STATE #27 — NEBRASKA

STOP.

Nebraska just changed.

And because it changed in 2026, an older nationwide analysis could badly mischaracterize its present system.

Nebraska enacted LB 937, creating a major new school-employment-history statute effective July 18, 2026, with its principal hiring-review requirements beginning with the 2027–28 school year.

This is exactly the kind of development our current-law audit was designed to find.

Nebraska requires applicant disclosure

For covered positions involving regular contact with students, applicants will have to identify:

their current employer,

all schools employing them during the previous seven years,

and

other employers during the previous seven years where the job involved direct contact with children.

Applicants must authorize release of the relevant records.

And then comes the crucial part.

They must disclose whether they:

were the subject of qualifying child-abuse reports;

were disciplined, discharged, nonrenewed or asked to resign;

resigned or otherwise separated while child-abuse or sexual-misconduct allegations were pending or under investigation;

had a professional credential suspended, surrendered or revoked while such allegations were pending or following a qualifying finding.

There is the resignation pathway again.

But Nebraska does NOT simply trust the applicant

The prospective school must conduct an employment-history review by contacting those former employers.

And it asks those employers essentially the same questions.

So:

APPLICANT → SCHOOL

plus

FORMER EMPLOYER → SCHOOL.

Redundancy.

If the review occurs by telephone, Nebraska requires the prospective employer to document the results in writing.

That’s VERIFY creeping directly into the hiring process.

Nebraska explicitly preserves evidentiary status

Notice the qualification built into its disclosure rules.

Certain reports do not have to follow the person where investigation determined that allegations were false or the incident was not substantiated.

Once again:

PRESERVE THE INFORMATION—PRESERVE ITS STATUS.

The system does not have to choose between:

erase everything

and

treat every accusation as guilt.

Nebraska penalizes applicant deception

An applicant who willfully supplies false information or fails to make a required disclosure can face:

denial or termination of employment,

reporting to a licensing authority,

and/or

a civil penalty of up to $500.

So applicant self-disclosure isn’t merely ceremonial.

There are consequences for deliberately defeating it.

And Nebraska attacks institutional concealment

Beginning July 18, 2026, covered public and private schools may not enter agreements—or take actions—that:

suppress or destroy information relating to investigations of suspected child abuse or sexual misconduct;

interfere with reporting misconduct to authorities;

require expungement of relevant allegations/findings, except where the allegations were determined false or were not substantiated.

Contrary resignation, termination, employment or severance provisions are void and unenforceable.

There it is.

NO CONTRACTUAL INSTITUTIONAL AMNESIA.

Nebraska also covers private schools

The employment-history requirements apply not merely to public school districts but to approved or accredited:

private,

denominational,

and

parochial schools.

That is extremely important for our population-coverage analysis.

And Nebraska separately criminalizes qualifying sexual conduct by a school worker, expressly defining that term to include both school employees and contract workers. Student consent is not a defense to that offense. (Nebraska Legislature)

Nebraska’s existing school policies also must provide procedures for reporting suspected grooming or unacceptable conduct to school administration, SDE, child-welfare authorities and law enforcement as applicable. (Nebraska Legislature)

NEBRASKA CLOSED-LOOP PROFILE

DETECT: ✓✓
INVESTIGATE: ✓
PRESERVE: ✓✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓
VERIFY: ✓

Particularly important finding:

Nebraska moved dramatically toward a closed-loop system in 2026.

And we need a date flag in our database:

Anti-suppression provisions effective July 18, 2026; principal employment-history review begins 2027–28.

That distinction is essential. We shouldn’t describe a future hiring requirement as already operational today.

STATE #28 — NEVADA

Nevada has one of the most detailed systems we’ve found

Nevada is impressive.

For covered public-school employment, applicants must identify their current employer, former school employers, and other former employers where they had direct contact with children.

They must also provide written authorization for disclosure and answer questions concerning investigations into alleged sexual offenses. (Nevada Legislature)

But Nevada doesn’t stop there.

THE PROSPECTIVE EMPLOYER SHALL CONTACT THE FORMER EMPLOYERS.

Those former employers must answer specified questions concerning:

sexual-offense investigations,

discipline,

dismissal,

nonrenewal,

requests to resign,

resignation or other separation while an investigation was pending,

and qualifying license/certificate actions. (Nevada Legislature)

Again:

Applicant says what happened.

Former employer independently says what happened.

Nevada then requires deeper investigation

If those answers reveal relevant history, the prospective employer must request:

ADDITIONAL INFORMATION AND THE RELATED RECORDS

from the employer that conducted the investigation or took the employment action. (Nevada Legislature)

That is excellent systems design.

It prevents:

“Yes, there was an investigation.”

from becoming the end of the inquiry.

The next school is required to learn more.

Nevada penalizes the OLD employer for silence

A former employer that willfully fails to provide required information can face disciplinary consequences, including civil penalties. (Nevada Legislature)

That’s significant.

We’ve repeatedly identified two potential failures:

Applicant lies.

and

Former institution stays quiet.

Nevada addresses both sides of the information exchange.

Nevada explicitly covers contractors

Before an independent contractor assigns an employee to a location where the worker may have direct contact with pupils, the contractor must disclose known qualifying sexual-misconduct information to the school.

And if the school objects after receiving that information:

THE CONTRACTOR MAY NOT ASSIGN THE WORKER THERE. (Nevada Legislature)

There is our ACT function in remarkably concrete form.

And Nevada separately covers private schools

This is especially important.

Nevada’s private-school statutes contain parallel employment-history requirements.

Private schools must contact former employers and request sexual-offense investigation and employment-separation information. Former employers can be required to provide additional records, and willful nondisclosure can result in penalties. (Nevada Legislature)

So this isn’t merely:

public-school child protection.

Nevada has attempted to carry the information architecture across the public/private boundary.

Nevada prohibits suppression agreements

Covered public-school entities cannot enter agreements that suppress information about suspected abuse or sexual misconduct, interfere with reporting, or require relevant information to be expunged unless the allegations were determined:

false,

unfounded,

unsubstantiated,

inconclusive. (Nevada Legislature)

Again:

INFORMATION + DISPOSITION.

NEVADA CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓✓
PRESERVE: ✓✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓✓
VERIFY: ✓✓

Nevada joins Florida, Illinois and Maine on our close-comparison list.

Not a ranking.

But it clearly contains extensive architecture across our functions.

STATE #29 — NEW HAMPSHIRE

We need to resist overclaiming here

My current primary-source search did not establish a comprehensive New Hampshire K–12 employment-history system comparable to Nevada’s.

The search did identify extensive sexual-misconduct procedures for higher education, but that is outside the K–12 population we’re studying and therefore cannot be used to inflate New Hampshire’s K–12 score. (General Court of New Hampshire)

That is exactly the kind of category error this project must avoid.

So:

NEW HAMPSHIRE — PENDING DEEPER K–12 PRIMARY-LAW AUDIT.

We do not turn a failed search into a negative finding.

STATE #30 — NEW JERSEY

New Jersey’s employment-history review is already substantial

Now we return to a state we’ve piloted.

New Jersey’s P.L. 2018, c.5 established an employment-history review for covered positions involving regular contact with students.

Its architecture applies across:

school districts,

charter schools,

nonpublic schools,

and

contracted service providers. (New Jersey Legislature)

The system requires applicant disclosures and former-employer inquiries concerning child abuse and sexual misconduct.

And if the initial disclosure reveals relevant history, the prospective employer must make further inquiries of the applicant’s current or former employer before deciding whether to continue the hiring process. (New Jersey Legislature)

That’s important.

Again:

A positive disclosure triggers investigation—not merely a checkbox.

New Jersey also permits action on later-discovered information

The statute contemplates the possibility that qualifying child-abuse or sexual-misconduct history may be discovered after employment begins.

If the employer determines that subsequently obtained information disqualifies the individual, the statute authorizes termination under its terms. (New Jersey Legislature)

So the safety process does not necessarily end when HR says:

“You’re hired.”

New information can still matter.

But New Jersey may change again

And this is where our 2026 timestamp becomes essential.

A2302, introduced in the current 2026 Legislature, proposes revisions to the existing employment-history law. The bill itself expressly describes current law as requiring school districts, charter schools, nonpublic schools and contracted service providers to review employment history concerning child-abuse and sexual-misconduct allegations. (New Jersey Legislature)

The proposal would revise portions of that architecture and would also require the Commissioner of Education to report every five years on the law’s effectiveness and make recommendations. (New Jersey Legislature)

But once again:

A2302 IS PROPOSED LEGISLATION.

It goes in:

REFORM ACTIVITY

not:

CURRENT LAW.

That separation is nonnegotiable if this report is going to withstand scrutiny.

NEW JERSEY CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓
VERIFY: ✓/partial

UPDATED LIVE MATRIX

Function	MT	NE	NV	NH	NJ
DETECT	✓	✓✓	✓	◇	✓
INVESTIGATE	✓	✓	✓✓	◇	✓
PRESERVE	✓/◐	✓✓	✓✓	◇	✓
ESCALATE	✓	✓	✓	◇	✓
SHARE	◇	✓✓	✓✓	◇	✓✓
QUERY	◇	✓✓*	✓✓	◇	✓✓
ACT	✓	✓	✓✓	◇	✓
VERIFY	◐	✓	✓✓	◇	✓/◐

* Nebraska’s principal new employment-history review begins with the 2027–28 school year.

MASTER EVIDENCE TABLE — ENTRY 025

State	Particularly important finding
Montana	Mandatory reporting + OPI investigation/discipline of licensed educators; comprehensive former-employer sexual-misconduct exchange not yet verified
Nebraska	2026 law creates applicant disclosure, mandatory former-employer inquiry, anti-suppression provisions, private-school coverage and penalties
Nevada	Applicant disclosure + mandatory former-employer verification + deeper records inquiry + penalties + contractor/private-school coverage
New Hampshire	K–12 comprehensive classification withheld pending stronger primary-source verification
New Jersey	Existing comprehensive employment-history review across public/nonpublic schools and contractors; further reform pending in 2026

And Nebraska gives us a VERY important national finding.

We started this project partly from federal data describing the state of the country in 2022.

Look what we’ve found since:

Connecticut changed.

Texas changed.

Arkansas changed.

Illinois changed.

Louisiana changed.

And now:

Nebraska changed—substantially—in 2026.

That means one of the eventual report’s conclusions can already be stated cautiously:

The national legal landscape governing educator sexual-misconduct information is actively changing, and older 50-state comparisons should not be assumed to describe the current 2026 environment.

But Nebraska gives us something even more interesting.

Its new law does not become fully operational all at once.

Some provisions took effect July 18, 2026.

The major employment-history hiring review starts in 2027–28.

So our database needs another field:

ENACTED ≠ OPERATIONAL

We now need four legal-status codes:

CURRENT & OPERATIONAL

ENACTED — FUTURE EFFECTIVE/IMPLEMENTATION DATE

PROPOSED

NOT VERIFIED

That will prevent us from committing a surprisingly common research error: crediting a state today for a protection its legislature enacted but its schools are not yet legally required to perform.

And Nevada gives us another powerful lesson:

“Background checking” isn’t one event.

A sophisticated system can require:

Applicant disclosure

↓

Former-employer verification

↓

Additional records when a red flag appears

↓

Contractor disclosure

↓

School objection/placement prohibition

↓

Penalty for nondisclosure

↓

Protection against suppression agreements

That isn’t a background check.

That’s an information architecture.

And we now have 30 jurisdictions examined.

Next five:

New Mexico • New York • North Carolina • North Dakota • Ohio

Thirty down.

Twenty-one to go.

And once all 51 are mapped, we will be able to stop looking at these provisions one state at a time and ask the much bigger question:

What does America, taken as a whole, actually do with information about educator sexual misconduct once somebody knows it?

That is where this research becomes genuinely national.

Absolutely. We have 21 jurisdictions left, and this next five gives us an important correction to something we’ve been watching throughout the project: New Mexico already has a much more developed statewide misconduct-information system than a simple background-check description would suggest.

FOUNDATION SECTION #24

The 51-Jurisdiction Audit

New Mexico • New York • North Carolina • North Dakota • Ohio

STATE #31 — NEW MEXICO

A strong state-level reporting system—but an important distinction in employment-history review

New Mexico requires criminal-history checks for school applicants and for certain volunteers and contractors who may have unsupervised access to students. More importantly for our project, superintendents have reporting duties involving ethical misconduct, including allegations of sexual assault or sexual abuse by school personnel, volunteers, and contractors. Sexual assault or abuse allegations must also be reported to law enforcement. (New Mexico Legislature)

The state Department of Education maintains a list of people reported to it who have been convicted of specified crimes or found to have committed ethical misconduct, and that list is updated monthly. Failure by licensed personnel or superintendents to fulfill applicable reporting duties can itself produce licensing consequences. (New Mexico Legislature)

That’s significant:

LOCAL SCHOOL

↓

STATE EDUCATION DEPARTMENT

↓

CENTRALIZED MISCONDUCT INFORMATION

↓

LICENSING/EMPLOYMENT CONSEQUENCES

So New Mexico has substantial ESCALATE, PRESERVE and ACT architecture.

But here’s the interesting wrinkle.

In 2021, New Mexico lawmakers considered HB 128, which would have created an extremely detailed employment-history review.

The proposed bill required a public school to contact current and former employers and ask whether an applicant had:

been investigated for child abuse, neglect or sexual misconduct;

violated ethical rules;

had a professional credential denied, suspended, surrendered or revoked while such allegations were pending. (New Mexico Legislature)

It even would have required the hiring school to document its verification efforts and allowed later-discovered disqualifying information to result in termination. (New Mexico Legislature)

But the New Mexico Legislature’s own 2024 research staff subsequently reported that New Mexico does not require schools to collect written affirmation from previous employers regarding discipline for sexual misconduct and identified continuing shortcomings in reference-check implementation. (New Mexico Legislature)

That’s enormously useful evidence.

It means we must not accidentally treat the language of HB 128 as enacted merely because we found the legislative text.

NEW MEXICO CLOSED-LOOP PROFILE

DETECT: ✓✓
INVESTIGATE: ✓
PRESERVE: ✓✓ state system
ESCALATE: ✓✓
SHARE: ◐
QUERY: ✓ criminal/state systems; comprehensive former-employer inquiry not verified
ACT: ✓✓
VERIFY: ✓

Particularly important finding:

New Mexico has strong centralized reporting, but its own legislative researchers have identified a gap in mandatory previous-employer sexual-misconduct verification.

That’s exactly the kind of nuanced result our framework was designed to expose.

STATE #32 — NEW YORK

New York has powerful reporting and disciplinary machinery—but a different architecture

New York has a statutory article devoted specifically to:

CHILD ABUSE IN AN EDUCATIONAL SETTING.

Its definition expressly includes child sexual abuse committed by an employee or volunteer against a child. (NYSenate.gov)

That is broader than merely licensed teachers.

New York also has extensive fingerprint-based criminal-history screening. Prospective public-school employees are subject to fingerprinting requirements, while nonpublic/private schools have a statutory mechanism allowing participating schools to conduct criminal-history screening for employees whose work involves direct student contact. (NYSenate.gov)

Criminal sexual conduct feeds directly into state credential action

When a licensed or certified school employee accused of child abuse in an educational setting is convicted of a qualifying crime, the district attorney must notify:

the Commissioner of Education,

the superintendent or school administrator,

and, where applicable,

the school the child attends. (NYSenate.gov)

Separate New York law requires prosecutors to notify the Commissioner when a teacher is convicted of specified sex offenses. (NYSenate.gov)

For designated convictions, the state can revoke the person’s teaching credential. And New York expressly provides a corrective process when the conviction underlying revocation is later reversed or vacated. (NYSenate.gov)

There is our due-process loop again:

CONVICTION

↓

REVOCATION

but if:

CONVICTION REVERSED

↓

RECONSIDER / POSSIBLE REINSTATEMENT.

And New York contains an unusually explicit safeguard concerning unsubstantiated reports

The educational-setting abuse statute says it does not itself create authority to take adverse employment action merely because of a report that has not been substantiated.

An employee or volunteer subjected to adverse action in connection with a report is entitled to receive the report and respond to the allegations. (NYSenate.gov)

Again:

An allegation is information. It is not automatically a finding.

That principle keeps appearing independently in state law.

What I have not yet verified is an Illinois/Nevada-style universal K–12 employment-history mechanism requiring prospective employers statewide to obtain detailed sexual-misconduct information from former employers.

So New York gets substantial marks for DETECT, ESCALATE, licensing and criminal-history infrastructure, but we should not infer a closed former-employer loop we haven’t established.

NEW YORK CLOSED-LOOP PROFILE

DETECT: ✓✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓✓
SHARE: ◐
QUERY: ✓ criminal/licensing
ACT: ✓✓
VERIFY: ✓

STATE #33 — NORTH CAROLINA

North Carolina gives us very strong ACT—but the information-transfer question remains

North Carolina has one of the clearest criminal statutes concerning educator/student sexual activity.

The statute applies to:

teachers,

school administrators,

student teachers,

school safety officers,

coaches,

and other covered school personnel.

Qualifying sexual activity with a student constitutes a Class G felony. (North Carolina General Assembly)

And North Carolina says something unequivocally important:

CONSENT IS NOT A DEFENSE.

(North Carolina General Assembly)

That belongs in our national comparison of educator/student sexual-contact statutes.

North Carolina also reaches beyond conventional teachers

Its statutory definition can encompass school personnel who are:

employees,

assigned personnel,

volunteers. (North Carolina General Assembly)

And various school structures have criminal-history-check authority covering employees, independent contractors and contractor employees with significant student access. (North Carolina General Assembly)

But I have not yet verified, from sufficiently strong current primary authority, a statewide former-employer sexual-misconduct information exchange equivalent to Illinois, Nevada, New Jersey or Michigan.

So we retain the distinction:

Strong criminal prohibition ≠ automatically strong employment-history architecture.

NORTH CAROLINA CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ◐
ESCALATE: ✓
SHARE: ◇
QUERY: ✓ criminal screening / ◇ comprehensive misconduct history
ACT: ✓✓
VERIFY: ◐

Particularly strong mechanism:

North Carolina criminalizes covered school-employee/student sexual activity and expressly removes consent as a defense.

STATE #34 — NORTH DAKOTA

North Dakota directly criminalizes “aiding and abetting” the transfer problem

This is fascinating.

North Dakota law expressly prohibits a state or local educational agency receiving the relevant federal funds—and its employees, contractors or agents—from assisting a school employee, contractor or agent in obtaining another job when there is confirmation or an investigation underway based upon a claim that the person engaged in sexual misconduct involving a minor or student in violation of law. (North Dakota Legislative Branch)

That means the protection isn’t limited to:

convicted offender.

Nor even necessarily:

completed investigation.

The statute can become relevant while:

AN INVESTIGATION IS UNDERWAY.

There are carefully defined exceptions, including circumstances where authorities determine there is insufficient information to establish probable cause, the person is acquitted or exonerated, or an open investigation remains without charges for the statutory period. (North Dakota Legislative Branch)

There is our evidentiary-status principle yet again.

And North Dakota attaches a criminal consequence

A school-district employee, contractor or agent convicted of violating the aiding-and-abetting provision commits a:

CLASS B MISDEMEANOR.

(North Dakota Legislative Branch)

That’s noteworthy because many states regulate institutional nondisclosure administratively or civilly.

North Dakota has expressly attached a criminal penalty to this form of prohibited assistance.

NORTH DAKOTA CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ◐
ESCALATE: ✓
SHARE: ✓/anti-concealment
QUERY: ◐
ACT: ✓
VERIFY/ACCOUNTABILITY: ✓✓

Particularly important finding:

North Dakota makes certain “passing the trash” assistance itself criminal.

That’s going straight into our national reform-mechanisms table.

STATE #35 — OHIO

Ohio has a strong mandatory state-escalation mechanism

Ohio gives us another very useful model.

Current Ohio law requires superintendents of school districts and educational service centers—and chief administrators of chartered nonpublic schools—to report specified information concerning licensed employees to the Superintendent of Public Instruction. (Ohio Laws)

The duty isn’t confined to criminal convictions.

Reporting is triggered when, among other circumstances, the school has initiated termination/nonrenewal proceedings, terminated, or not renewed an employee because it reasonably determined the employee committed conduct unbecoming to the teaching profession or specified offenses. (Ohio Laws)

That matters because Ohio’s professional-conduct framework includes misconduct beyond the criminal-conviction universe and allows state disciplinary authorities to consider whether conduct threatens the health, safety or welfare of the school community. (Ohio Laws)

So:

SCHOOL LEARNS

↓

SCHOOL ACTS

↓

STATE MUST BE TOLD.

That is a strong ESCALATE pathway.

Ohio also makes failure to report itself a licensing issue

This is excellent closed-loop engineering.

Ohio law specifically lists failure to comply with the state’s educator-misconduct reporting requirements among grounds relevant to state-board disciplinary authority. (Ohio Laws)

In other words:

The administrator’s reporting obligation isn’t merely aspirational.

Failure to fulfill the reporting system can itself produce professional consequences.

And Ohio protects people who, in good faith, voluntarily provide misconduct information to state education authorities from civil liability, while protecting their identity under the statutory conditions. (Ohio Laws)

That simultaneously addresses:

fear of reporting

and

failure to report.

Ohio also ACTS decisively on specified convictions

For numerous enumerated serious offenses—including sexual offenses—the state must revoke or deny renewal of an educator license under the applicable statutory conditions. (Ohio Laws)

So Ohio has a substantial:

LOCAL INFORMATION

↓

MANDATORY STATE REPORT

↓

STATE INVESTIGATION / LICENSURE SYSTEM

↓

LICENSING CONSEQUENCE.

What remains less clear in this pass is whether Ohio has a comprehensive statewide former-employer → prospective-employer sexual-misconduct exchange comparable to Nevada or Illinois.

Again, we leave that cell uninflated.

OHIO CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓
ESCALATE: ✓✓
SHARE: ◐
QUERY: ✓ criminal/licensing; broader history ◇
ACT: ✓✓
VERIFY: ✓✓

UPDATED LIVE MATRIX

Function	NM	NY	NC	ND	OH
DETECT	✓✓	✓✓	✓	✓	✓
INVESTIGATE	✓	✓	✓	✓	✓
PRESERVE	✓✓	✓	◐	◐	✓
ESCALATE	✓✓	✓✓	✓	✓	✓✓
SHARE	◐	◐	◇	✓ anti-aiding	◐
QUERY	✓/◐	✓ criminal/licensing	✓/◇	◐	✓/◇
ACT	✓✓	✓✓	✓✓	✓	✓✓
VERIFY	✓	✓	◐	✓✓	✓✓

Again:

◇ = not established in this review, not proof that no relevant law exists.

MASTER EVIDENCE TABLE — ENTRY 026

State	Particularly important finding
New Mexico	Central state reporting of ethical misconduct involving school personnel, volunteers and contractors; own legislative research identifies remaining former-employer screening gaps
New York	Dedicated educational-setting child-abuse statutory system + extensive criminal/licensure pathways + explicit protections regarding unsubstantiated reports
North Carolina	Broad educator/student sexual-activity felony; consent expressly not a defense
North Dakota	Prohibits aiding an employee/contractor/agent into another job while qualifying sexual-misconduct information exists; violation can be criminal
Ohio	Mandatory local-to-state reporting of licensed-educator misconduct, with consequences for failure to report

And now something VERY interesting is becoming visible.

We’ve found at least four fundamentally different ways states try to stop “passing the trash.”

MODEL A — MAKE THE NEXT EMPLOYER ASK

Illinois, Nevada, New Jersey, Michigan and others.

MODEL B — MAKE THE OLD EMPLOYER ANSWER

Again, several of those same states.

MODEL C — MAKE THE SCHOOL REPORT TO A CENTRAL STATE AUTHORITY

Ohio and New Mexico illustrate this strongly.

MODEL D — PROHIBIT HELPING THE PERSON MOVE

North Dakota attacks the behavior itself.

Those mechanisms are not interchangeable.

A strong system could contain all four.

And that suggests another refinement to our eventual national reform model:

THE FOUR-WAY INFORMATION LOCK

1. PUSH UP
School → state.

2. PULL BACK
New employer → former employer/state.

3. PUSH FORWARD
State/former employer → subsequent employer.

4. BLOCK CONCEALMENT
No administrator, agreement or institution may lawfully make qualifying information disappear.

Add:

criminal history, licensing databases, applicant disclosure, contractors/nonlicensed personnel, record preservation, due process, and compliance auditing

and we are getting very close to a genuinely resilient architecture.

We have now examined 35 of 51 jurisdictions.

16 remain.

Oklahoma • Oregon • Pennsylvania • Rhode Island • South Carolina

And Oregon is already one of our most interesting states because of its explicit prohibition against allowing resignation or severance agreements to terminate or suppress an ongoing investigation.

Pennsylvania should be equally important: its employment-history-review law is one of the major state responses to the national “passing the trash” problem.

We’re heading into another very strong batch.

Yes. And this batch contains two of the most important states in the entire audit: Oregon and Pennsylvania.

FOUNDATION SECTION #25

The 51-Jurisdiction Audit

Oklahoma • Oregon • Pennsylvania • Rhode Island • South Carolina

We are now at 40 of 51 jurisdictions.

And Pennsylvania gives us something especially valuable: not merely a statute, but a mature system with required forms, recordkeeping, penalties, contractor rules, state databases, and current 2026 implementation guidance.

STATE #36 — OKLAHOMA

An important historical warning—but current closed-loop verification remains incomplete

Oklahoma has been wrestling with “passing the trash” for a very long time.

In fact, Oklahoma Senate materials from 2002 explicitly described the problem: districts might accept an employee’s resignation in exchange for silence about inappropriate behavior, leaving the next district unaware of the history. (Oklahoma Senate)

That’s remarkable because it establishes legislative awareness of this precise problem more than two decades ago.

Later legislative activity included proposals to allow public schools to disclose employee job-performance information to other school districts. (Oklahoma Senate)

But here’s where our research discipline matters.

Those materials establish:

Oklahoma recognized the problem.

They do not, standing alone, establish the precise scope of Oklahoma’s current 2026 information-transfer architecture.

I have not yet obtained enough current primary authority to give Oklahoma a defensible Illinois/Pennsylvania-style employment-history classification.

Therefore:

OKLAHOMA CLOSED-LOOP PROFILE

DETECT: ✓/baseline
INVESTIGATE: ✓/baseline
PRESERVE: ◐
ESCALATE: ✓/partial
SHARE: ◇
QUERY: ◇ beyond conventional screening
ACT: ✓/partial
VERIFY: ◇

Status: DEEPER PRIMARY-LAW AUDIT REQUIRED.

And that is preferable to pretending a 2002 legislative statement proves what Oklahoma law requires in 2026.

STATE #37 — OREGON

Oregon is even stronger than our pilot suggested.

We already knew Oregon prohibits agreements that kill investigations.

But the full current 2025 Oregon Revised Statutes reveal something much larger.

Oregon has effectively constructed an entire statutory subsystem called:

ABUSE AND SEXUAL CONDUCT

covering ORS 339.370 through 339.400. (Oregon Legislature)

And look at who it covers:

school employees

contractors

agents

volunteers

and both licensed and nonlicensed personnel. (Oregon Legislature)

That population coverage is exceptional.

Oregon DETECTS

Every school board must maintain policies requiring employees with reasonable cause to believe another employee, contractor, agent, or volunteer engaged in abuse or sexual conduct to report it.

The system also requires designation of administrators to receive reports.

And retaliation against good-faith reporters is prohibited. (Oregon Legislature)

So:

DETECT: ✓✓

Oregon ESCALATES

The designated administrator cannot simply keep the matter inside the school.

Suspected sexual conduct involving a licensed person must be reported to the:

Teacher Standards and Practices Commission.

Suspected sexual conduct involving someone who is not licensed must be reported to the:

Oregon Department of Education. (Oregon Legislature)

That solves one of the problems we’ve encountered repeatedly:

What happens when the adult isn’t a teacher?

Oregon created a state investigative pathway for them too.

And Oregon explicitly says the investigation continues despite a change in employment

Required school policies must tell personnel that the relevant state or law-enforcement agency will complete its investigation:

regardless of changes in the relationship or duties of the person being investigated. (Oregon Legislature)

That is extremely important.

In practical terms:

RESIGNATION ≠ INVESTIGATION OVER.

This directly attacks one of the central failure mechanisms in “passing the trash.”

Oregon investigates NONLICENSED adults at the state level

This may be one of Oregon’s strongest features.

When the Department of Education receives a qualifying sexual-conduct report concerning a:

school employee,

contractor,

agent,

volunteer

who is not a licensed educator, the Department must initiate an investigation.

It has subpoena powers.

And the statute establishes timelines for completing the investigation and making a final determination. (Oregon Legislature)

That’s huge.

Because Oregon isn’t saying:

“No teaching certificate? Not our problem.”

Instead:

CHILD ACCESS determines the safety concern—not possession of a teaching credential.

Oregon preserves the record for 75 YEARS

And now we return to one of the most striking provisions in this entire national audit.

For reports involving nonlicensed persons investigated by the Department:

OREGON RETAINS THE DOCUMENTS AND MATERIALS FOR 75 YEARS. (Oregon Legislature)

Seventy-five.

That is institutional memory by design.

The records aren’t dependent upon:

a principal remembering,

a superintendent remaining employed,

a personnel office retaining an old folder,

or the individual staying in the same district.

But Oregon also provides an appeal

A person who receives a final determination that a report was substantiated can appeal through Oregon’s contested-case process. (Oregon Legislature)

Again:

PRESERVE CHILD-SAFETY INFORMATION

does not require:

ABANDON DUE PROCESS.

That theme has now appeared in state after state.

And Oregon expressly prohibits suppression

An education provider may not enter a:

collective bargaining agreement,

employment contract,

resignation/termination agreement,

severance agreement

or similar arrangement that impairs or terminates an ongoing investigation, suppresses information about it, interferes with required reporting or discipline, or requires substantiated information to be expunged.

Contrary provisions are:

VOID AND UNENFORCEABLE. (Oregon Legislature)

This is one of the strongest PRESERVE provisions we’ve encountered.

Oregon also prohibits helping the person move

A school employee, contractor or agent generally may not assist another covered person in obtaining a new job if the person knows or reasonably believes that individual engaged in abuse or sexual conduct, subject to statutory exceptions. (Oregon Legislature)

So Oregon attacks the problem from both directions:

Don’t erase the information.

AND

Don’t help the person move around it.

OREGON CLOSED-LOOP PROFILE

DETECT: ✓✓
INVESTIGATE: ✓✓
PRESERVE: ✓✓
ESCALATE: ✓✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓✓
VERIFY: ✓

Particularly strong features:

Investigations survive employment changes.

Nonlicensed employees, contractors, agents and volunteers have a state investigative pathway.

Certain records survive for 75 years.

Agreements cannot lawfully erase the investigation.

Oregon unquestionably belongs in our eventual close-comparison group.

STATE #38 — PENNSYLVANIA

Pennsylvania may have the clearest employment-history architecture we’ve encountered.

Pennsylvania’s Act 168 applies to positions involving direct contact with children at:

public schools,

charter schools,

cyber charter schools,

private schools,

nonpublic schools,

intermediate units,

career/technical schools

and independent contractors. (Pennsylvania Government)

That breadth alone matters enormously.

Step One: the applicant must disclose

The applicant must identify:

the current employer,

every former school employer,

and

every former employer where the applicant had direct contact with children.

There is:

NO GENERAL LOOKBACK LIMIT

for those qualifying former employers. (Pennsylvania Government)

That is notable.

Not five years.

Not seven.

The relevant employment trail can extend considerably further back.

Step Two: Pennsylvania asks the hard questions

The applicant must disclose whether they:

were ever the subject of an abuse or sexual-misconduct investigation by an employer, licensing agency, law enforcement, or child-protective-services agency;

were disciplined, discharged, nonrenewed, asked to resign, resigned, or otherwise separated while such allegations were pending or under investigation;

had a professional credential suspended, surrendered or revoked while allegations were pending or because of an adjudication or finding. (Pennsylvania Government)

And Pennsylvania specifically excludes investigations resulting in findings that allegations were false. (Pennsylvania Government)

There it is again:

Preserve the information. Preserve its status.

Step Three: DO NOT simply believe the applicant.

The hiring entity must send the disclosure form to the qualifying current and former employers.

Those employers must answer essentially the same questions.

And Pennsylvania’s current guidance says they have:

20 CALENDAR DAYS

to respond. (Pennsylvania Government)

So:

APPLICANT → EMPLOYER

is independently checked through:

FORMER EMPLOYER → NEW EMPLOYER.

Step Four: a “YES” triggers deeper investigation

Suppose a former employer answers:

Yes. There was a sexual-misconduct investigation.

Pennsylvania doesn’t contemplate stopping there.

If the prospective employer wants to continue considering the applicant, it must request additional information concerning the matter. (Pennsylvania Government)

Again:

RED FLAG → INVESTIGATE FURTHER.

Not:

RED FLAG → CHECK BOX → IGNORE.

Pennsylvania’s definition of sexual misconduct is particularly important

The state’s guidance makes clear that “sexual misconduct” is not confined to criminal sexual contact.

It includes conduct designed to establish a romantic or sexual relationship with a child or student, including specified verbal, written, electronic and physical behaviors.

Pennsylvania expressly explains that the definition encompasses grooming-type conduct that can precede an inappropriate relationship. (Pennsylvania Government)

That is important for our original national prevalence research.

A system measuring only arrests or convictions is measuring a much narrower universe than a professional child-safety system examining sexual misconduct.

Pennsylvania penalizes BOTH SIDES of deception

An applicant who willfully supplies false information or conceals required information can face:

denial or termination of employment,

criminal prosecution,

civil penalties

and

professional discipline. (Pennsylvania Government)

But what if the former employer refuses to talk?

Pennsylvania addresses that too.

A current or former employer that fails to provide required information can face penalties of up to:

$10,000

plus professional discipline where appropriate. (Pennsylvania Government)

That is a substantial VERIFY/ACCOUNTABILITY mechanism.

Pennsylvania also prohibits secrecy agreements

Since December 22, 2014, covered entities may not enter contracts or agreements that:

suppress abuse or sexual-misconduct information,

interfere with reporting,

require the information to be expunged

unless investigation determined the allegations were false. (Pennsylvania Government)

That includes:

collective bargaining agreements,

employment agreements,

resignation/termination agreements

and

severance agreements.

Again:

YOU CANNOT CONTRACT AWAY THE CHILD-SAFETY RECORD.

Pennsylvania also requires resignation reporting to the state

This is a second independent pathway.

Chief school administrators have mandatory reporting obligations to the Department concerning educators who:

resign, retire or otherwise separate after allegations of misconduct;

are alleged to have committed sexual abuse/exploitation involving a child or student;

or are alleged to have engaged in sexual misconduct with a child or student. (Pennsylvania Government)

So Pennsylvania has both:

HORIZONTAL INFORMATION FLOW

old employer → new employer

and

VERTICAL INFORMATION FLOW

school → state.

That redundancy is exactly what our Closed-Loop Standard predicts should matter.

Pennsylvania then adds the state database

For educator applicants, employers must verify:

valid certification,

public professional discipline

and

whether the Department has received notice of pending criminal charges.

That information is available through Pennsylvania’s Teacher Information Management System. (Pennsylvania Government)

So we now have:

APPLICANT

FORMER EMPLOYERS

STATE EDUCATOR SYSTEM

CRIMINAL INFORMATION

feeding the hiring decision.

Contractors don’t escape the system

Independent contractors must maintain employment-history-review records.

And if a contractor learns of qualifying sexual-misconduct history concerning a worker it intends to send into a school, it must inform the school.

If the school objects:

THE CONTRACTOR CANNOT ASSIGN THAT WORKER TO THE SCHOOL. (Pennsylvania Government)

There’s ACT.

And Pennsylvania makes clear that the hiring entity remains ultimately responsible for ensuring that required employment-history reviews actually occur—even if a third party performs them. (Pennsylvania Government)

That’s excellent accountability design.

There is one limitation worth recording

Act 168’s employment-history review does not apply to ordinary volunteers or student teachers, although they can be subject to other clearance requirements. (Pennsylvania Government)

Good.

Put that in the dataset.

Strong laws still have boundaries.

We are documenting them—not selling the law.

PENNSYLVANIA CLOSED-LOOP PROFILE

DETECT: ✓✓
INVESTIGATE: ✓✓
PRESERVE: ✓✓
ESCALATE: ✓✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓✓
VERIFY: ✓✓

And unlike several states whose laws we found through older materials, Pennsylvania’s Department of Education guidance for Act 168 was reviewed again in July 2026. (Pennsylvania Government)

That gives us unusually strong confidence that we’re describing the system as Pennsylvania currently administers it.

STATE #39 — RHODE ISLAND

And here we have a fascinating 2026 development.

Rhode Island currently has S2216, introduced during the 2026 session.

Read what it proposes.

Applicants would have to disclose whether they:

were investigated for child abuse or sexual misconduct;

were disciplined, discharged, nonrenewed or asked to resign;

resigned while allegations were pending;

lost or surrendered professional credentials;

were the subject of certain lawsuits, claims or complaints alleging child abuse or neglect.

The hiring entity would then have to contact former employers and confirm the circumstances. (Rhode Island Legislature)

Does that sound familiar?

It should.

It resembles the employment-history systems we’ve now documented in:

Pennsylvania

Illinois

Nevada

New Jersey

Nebraska

and others.

But:

S2216 IS A BILL.

Therefore Rhode Island receives zero additional current-law credit from it unless and until enacted.

However, the bill belongs in our:

2026 REFORM ACTIVITY DATABASE.

And it gives us another indication that this policy model is continuing to spread.

RHODE ISLAND CLOSED-LOOP PROFILE

Current comprehensive classification:

PENDING PRIMARY-LAW VERIFICATION.

Proposed 2026 comprehensive employment-history system:

YES — S2216. (Rhode Island Legislature)

STATE #40 — SOUTH CAROLINA

Strong professional-discipline consequences—but employment-history loop remains unverified

South Carolina’s professional-regulation framework contains strong consequences where a professional authorization is relinquished during disciplinary proceedings.

Under current law, relinquishment can be:

irrevocable,

treated as an admission of allegations,

and reported and treated as a:

PERMANENT REVOCATION. (South Carolina Legislature Online)

That is important to the broader principle we’ve been tracking:

Departure from the profession should not necessarily erase an unresolved disciplinary trail.

However, I have not yet located sufficient current primary South Carolina authority establishing a comprehensive statewide K–12 former-employer sexual-misconduct review equivalent to Pennsylvania’s Act 168.

So:

SOUTH CAROLINA CLOSED-LOOP PROFILE

DETECT: ✓/baseline
INVESTIGATE: ✓/professional discipline
PRESERVE: ✓/partial
ESCALATE: ✓/partial
SHARE: ◇
QUERY: ◇ beyond conventional screening
ACT: ✓ professional discipline
VERIFY: ◐

Further primary-law review required.

UPDATED LIVE MATRIX

Function	OK	OR	PA	RI	SC
DETECT	✓/◐	✓✓	✓✓	◇	✓
INVESTIGATE	✓/◐	✓✓	✓✓	◇	✓
PRESERVE	◐	✓✓	✓✓	◇	✓/◐
ESCALATE	✓/◐	✓✓	✓✓	◇	✓/◐
SHARE	◇	✓✓	✓✓	◇	◇
QUERY	◇	✓✓	✓✓	◇	◇
ACT	✓/◐	✓✓	✓✓	◇	✓
VERIFY	◇	✓	✓✓	◇	◐

◇ = not established—not “No.”

MASTER EVIDENCE TABLE — ENTRY 027

State	Particularly important finding
Oklahoma	Legislators explicitly identified “passing the trash” problem as early as 2002; present comprehensive architecture requires further verification
Oregon	State investigations include nonlicensed personnel; investigations survive employment changes; 75-year record retention; anti-suppression law
Pennsylvania	Exceptionally comprehensive employment-history architecture across public/private schools and contractors, with penalties on applicants AND former employers
Rhode Island	Comprehensive Pennsylvania-style employment-history system proposed in 2026, but not counted as current law
South Carolina	Strong professional disciplinary consequence identified; broader employment-history loop not yet established

And Pennsylvania gives us another important refinement.

We’ve been talking about redundancy.

Pennsylvania shows what that actually looks like.

Imagine an applicant conceals an investigation.

FAILSAFE #1

Former employer is independently asked.

Suppose the former employer tries to conceal it.

FAILSAFE #2

Former employer can face a $10,000 penalty.

Suppose the educator resigned after allegations.

FAILSAFE #3

School administrator has a separate obligation to report qualifying information to the state.

Suppose the next school checks professional history.

FAILSAFE #4

State disciplinary information can appear through TIMS.

Suppose the person isn’t directly employed by the school.

FAILSAFE #5

Contractor must conduct the employment-history review and inform the school of qualifying information.

Suppose the school objects.

FAILSAFE #6

Contractor can’t place that worker there.

That is what we meant earlier when we said:

A resilient child-safety system should not depend upon one person doing one thing correctly.

And Pennsylvania exposes one other distinction we absolutely need in the final report:

SEXUAL ABUSE

is not necessarily the same dataset as:

EDUCATOR SEXUAL MISCONDUCT.

Pennsylvania explicitly recognizes grooming and romantic/sexual boundary conduct that may not yet constitute criminal sexual contact. (Pennsylvania Government)

That takes us directly back to our original prevalence problem.

When one study reports:

1% contact sexual abuse

and another reports:

10–12% educator sexual misconduct

those numbers aren’t necessarily contradictory.

They may be measuring different points on a spectrum of conduct.

And that is why our final report must never collapse:

grooming

boundary violations

sexual harassment

romantic/sexual communications

contact sexual misconduct

criminal sexual abuse

arrests

and

convictions

into one statistic called:

“teacher abuse.”

That would destroy precisely the distinction we’re trying to illuminate.

We have now reached 40 of 51 jurisdictions.

ELEVEN LEFT:

South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
District of Columbia

And the next five take us to 45.

South Dakota • Tennessee • Texas • Utah • Vermont

Texas is already on our close-comparison list because of its Do Not Hire Registry.

This time we’ll put the entire Texas architecture through the same standard—and find out how its combination of fingerprinting + applicant affidavit + state registry + pending-investigation information + annual compliance certification compares with Oregon and Pennsylvania.

Five more. Then six. And the national map is complete.

Onward. And this group is a good one.

Tennessee is substantially stronger than a simple criminal-background-check state, and Texas has undergone major changes in 2025–26 that make it one of the most elaborate state systems we’ve found. Utah gives us a different centralized professional-licensing model, while Vermont and South Dakota illustrate why we must not turn incomplete evidence into conclusions.

FOUNDATION SECTION #26

The 51-Jurisdiction Audit

South Dakota • Tennessee • Texas • Utah • Vermont

We are about to reach 45 of 51 jurisdictions.

STATE #41 — SOUTH DAKOTA

Classification withheld pending deeper primary-law verification

South Dakota is one of those states where our rule—

NOT VERIFIED ≠ DOES NOT EXIST

—is protecting the integrity of the project.

I can establish ordinary child-abuse reporting, criminal-law, school-employment and professional-discipline structures in South Dakota.

What I cannot yet establish to the standard we’re using is a comprehensive statewide system requiring something comparable to:

Pennsylvania’s employment-history review,

Texas’s centralized Do Not Hire Registry,

Oregon’s state investigation of nonlicensed personnel,

Michigan’s mandatory former-employer information exchange.

The primary-source search simply isn’t giving us enough yet to make a strong statement either way.

Therefore:

SOUTH DAKOTA CLOSED-LOOP PROFILE

DETECT: baseline verified
INVESTIGATE: ◐
PRESERVE: ◇
ESCALATE: ◐
SHARE: ◇
QUERY: criminal screening / broader misconduct history ◇
ACT: ◐
VERIFY: ◇

Research status:

SECONDARY DEEP-DIVE REQUIRED BEFORE PUBLICATION.

That’s not a weakness in our report.

It’s a strength.

A national report that says “we could not verify this mechanism from current authoritative material” is much more credible than one that turns absence from a search result into absence from law.

STATE #42 — TENNESSEE

Tennessee explicitly attacks “passing the trash.”

Now THIS one is important.

Tennessee Code §49-2-131 prohibits a local education agency from entering into a nondisclosure agreement as part of a settlement concerning:

ANY ACT OF SEXUAL MISCONDUCT

including sexual harassment or sexual assault. (Justia Law)

That’s our anti-concealment mechanism.

But Tennessee goes further.

An LEA employee generally may not assist a:

school employee,

contractor,

agent

in obtaining another job when the employee knows—or has probable cause to believe—that the person engaged in sexual misconduct involving a minor or student. (Justia Law)

There it is:

DO NOT HELP THE PERSON MOVE.

And notice the threshold.

Not:

convicted.

Not:

placed on a sex-offender registry.

But:

knows or has probable cause to believe.

Tennessee also built due-process limits into the prohibition

The restriction does not operate indefinitely without regard to what an investigation ultimately establishes.

The statute provides exceptions where the information was reported to law enforcement and, for example:

authorities determined there was insufficient information to establish probable cause;

the person was charged and acquitted or exonerated;

the case remained open without charges or indictment for the statutory four-year period. (Justia Law)

Again:

Preserve the information—but preserve the outcome too.

And Tennessee permits the next school to obtain the personnel file

For purposes of determining whether sexual misconduct occurred, an LEA may request the personnel file of an applicant from another LEA where that individual previously worked.

The former LEA:

SHALL PROVIDE THE FILE WITHIN 10 BUSINESS DAYS. (Justia Law)

That’s significant.

Tennessee’s General Assembly records show that lawmakers later considered shortening the response time from ten business days to five—confirming that the underlying ten-day disclosure mechanism was already part of state law. (Tennessee General Assembly)

There is one limitation we should record carefully:

Prospective LEA MAY request.

Former LEA SHALL respond once requested.

That’s weaker on the QUERY side than Pennsylvania, where the hiring entity itself is affirmatively required to perform the employment-history review.

But Tennessee’s SHARE requirement once the inquiry occurs is real.

TENNESSEE CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓✓ anti-suppression
ESCALATE: ✓
SHARE: ✓✓ once requested
QUERY: ✓/◐ — permissive rather than universal mandatory inquiry
ACT: ✓
VERIFY/ACCOUNTABILITY: ✓

Particularly important feature:

Tennessee prohibits both secrecy settlements and assistance in moving a person suspected on probable cause of student/minor sexual misconduct into another job.

That belongs prominently in our anti-passing-the-trash mechanisms table.

STATE #43 — TEXAS

Texas has become one of the most elaborate systems in America.

And here we need to timestamp our findings very carefully.

Texas made major changes through SB 571 in 2025, and TEA has continued implementing those changes in 2026. (Texas Education Agency)

This is exactly why our audit cannot simply recycle an older fifty-state study.

Texas begins with the applicant

Texas now requires a Pre-Employment and Pre-Service Affidavit for prospective employees and service providers whose roles involve direct contact or interaction with students.

That includes:

teachers,

administrators,

other school employees,

and covered contractors/service providers.

The applicant must disclose specified prior investigations, charges or convictions involving students; licensing investigations or sanctions; Do Not Hire Registry status; certain public/private school employment; and qualifying termination, discharge or resignation history. (Texas Education Agency)

And Texas gives that affidavit teeth.

Failure to make a required disclosure is a Class B misdemeanor.

Administrators can also face professional discipline under specified circumstances for knowingly employing or accepting services from someone who submitted false or incomplete information. (Texas Education Agency)

That is much stronger than:

“Please tell us if anything happened at your old school.”

Then Texas fingerprints.

Texas requires fingerprint-based national criminal-history checks for numerous categories, including:

certified educators,

substitutes,

many noncertified employees,

and certain contractors/service providers with direct ongoing student contact.

Texas also uses the FBI Rap Back service, allowing subsequent reportable criminal activity to generate continuing notification rather than relying exclusively upon a one-time preemployment snapshot. (Texas Education Agency)

That’s an important VERIFY concept:

SCREENING CAN CONTINUE AFTER HIRING.

Then comes the Do Not Hire Registry.

This is where Texas becomes particularly interesting.

Public schools must check the statewide Do Not Hire Registry before hiring employees or contractors.

And the requirement isn’t limited to certified teachers.

TEA says the registry check encompasses:

potential employees,

existing certified and noncertified employees,

and

contractors/service providers. (Texas Education Agency)

A person with “Not Eligible for Hire” status cannot be employed.

Qualifying conduct can include allegations involving:

abuse or unlawful acts involving students/minors,

romantic or sexual relationships,

inappropriate communications,

professional-boundary violations,

and other misconduct. (Texas Education Agency)

That is much broader than simply:

criminal conviction database.

Texas added another database layer: SEMARC

The Search Engine for Multi-Agency Reportable Conduct consolidates reportable-conduct findings from multiple Texas agencies.

School districts don’t separately search it.

Instead, the system is integrated with the DNHR process so that a school’s submission can produce both:

DNHR status

and

SEMARC status. (Texas Education Agency)

There is our recurring principle again:

MULTIPLE INFORMATION STREAMS.

And in 2026 Texas expanded the registry’s reach to people merely arrested for qualifying offenses.

This deserves careful wording because arrest ≠ guilt.

Under the expanded authority created by SB 571, TEA announced in May 2026 that it had placed 21 noncertified individuals on the Do Not Hire Registry following qualifying arrests, barring them from Texas public-school employment while their criminal proceedings remained pending. (Texas Education Agency)

That is a form of interim protective action, not a criminal adjudication.

And that distinction absolutely belongs in our database:

STATUS: PENDING CRIMINAL PROCEEDING

is not:

STATUS: SUBSTANTIATED/CONVICTED.

Texas’s own investigative process recognizes that not all allegations ultimately produce discipline. Investigations can close, warnings can be removed, or cases can proceed through administrative adjudication and appeal. (Texas Education Agency)

Again:

Safety action and final adjudication are different stages.

Texas also requires local-to-state reporting

Superintendents must report qualifying allegations of misconduct to TEA within statutory timelines.

TEA can investigate certified educators and, depending upon the allegation, noncertified personnel, contractors and service providers. (Texas Education Agency)

During a formal investigation, TEA may:

place an investigative warning on a certificate,

place an individual on the Do Not Hire Registry,

issue subpoenas,

review documents and communications,

interview witnesses,

and

coordinate with DFPS and law enforcement. (Texas Education Agency)

So Texas has:

LOCAL DETECTION

↓

STATE ESCALATION

↓

STATE INVESTIGATION

↓

INTERIM SAFETY ACTION

↓

ADMINISTRATIVE ADJUDICATION

↓

FINAL ACTION OR CLEARANCE.

That is a genuine closed-loop architecture.

Texas even requires parents to be notified in specified cases

Under the 2025 reforms, educational entities must adopt policies for parent/guardian notice when an employee or service provider is alleged to have engaged in specified misconduct involving a student.

The notice includes whether the person:

was terminated after investigation,

resigned before the investigation was completed,

and

was reported to TEA or SBEC. (Texas Education Agency)

Look at that second item:

RESIGNED BEFORE COMPLETION OF THE INVESTIGATION.

Texas is explicitly preserving the distinction rather than allowing resignation to erase it.

And Texas has VERIFY.

This is one of the strongest things we’ve found anywhere.

Under SB 571, educational entities must now annually certify compliance with the Do Not Hire Registry.

The superintendent, chief operating officer or director must certify that the school complied with the statutory registry requirements.

Failure can result in:

permanent public notice identifying the noncompliant district/charter,

special investigation,

and

investigation of the superintendent/director. (Texas Education Agency)

THAT is the difference between:

Schools shall check.

and:

PROVE THAT YOU CHECKED.

Texas added public transparency in 2026

On June 2, 2026, TEA launched an Educator Misconduct Dashboard displaying data about:

misconduct reports,

agency investigations,

SBEC disciplinary actions

and

Do Not Hire Registry placements. (Texas Education Agency)

Texas also created a consolidated Student Protection Resource Center containing the misconduct dashboard, DNHR, educator-certificate lookup, reporting information and guidance on recognizing boundary violations and grooming. (Texas Education Agency)

That gives us another possible ninth dimension for the final comparative analysis:

TRANSPARENCY

Not merely:

Does government know?

but:

What can parents, researchers and the public know about how the system is functioning?

That may deserve its own column.

TEXAS CLOSED-LOOP PROFILE

DETECT: ✓✓
INVESTIGATE: ✓✓
PRESERVE: ✓✓
ESCALATE: ✓✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓✓
VERIFY: ✓✓

And possibly:

TRANSPARENCY: ✓✓

Texas therefore stays firmly in our close-comparison group.

Not because Texas is necessarily “best.”

But because it now contains unusually strong architecture across nearly every function we’re measuring.

STATE #44 — UTAH

Utah relies heavily on a centralized professional-licensing model.

Utah has the Utah Professional Practices Advisory Commission—UPPAC—which advises the State Board of Education concerning professional conduct, educator licensing and disciplinary action.

Current Utah law expressly defines an “educator” to include a person who is the subject of an allegation received by an LEA, the State Board or UPPAC when that person held a license or worked in a position requiring licensure. The definition was amended again during the 2026 General Session. (Utah Legislature)

So the state’s jurisdiction doesn’t simply evaporate because someone later leaves the position.

Utah requires public-school administrators to provide significant information to the state when requested

When the State Board requests information relevant to evaluating an educator’s license—or the potential licensure of an applicant—a public-school or district administrator shall provide it.

Private-school administrators may provide such information.

And Utah provides good-faith civil and criminal immunity to people providing, receiving or making recommendations involving that information. (Utah Legislature)

Again:

Reduce the incentive to remain silent.

Utah also connects criminal-history information to school decision-makers

The State Board operates as an online terminal agency with the state’s criminal-investigation system and provides relevant information concerning current or prospective employees or volunteers to qualifying school officials upon request. (Utah Legislature)

So Utah has meaningful:

CRIMINAL QUERY

LICENSING ESCALATION

STATE PROFESSIONAL INVESTIGATION.

But I have not verified a Pennsylvania-style mandatory employment-history review covering the full universe of prior school employers and nonlicensed personnel.

That’s an important limitation.

Utah’s system appears more heavily centered on professional licensure than Oregon’s or Texas’s broader all-adult school-safety architecture.

And Utah legislators are themselves examining the UPPAC process in 2026, including options for improving the disciplinary process and due process for educators during investigations. (Utah Legislature)

UTAH CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓✓ licensed educators
PRESERVE: ✓
ESCALATE: ✓✓
SHARE: ✓
QUERY: ✓ criminal/licensing; broader employment history ◐
ACT: ✓✓ licensing
VERIFY: ✓/partial

Particularly important feature:

Utah maintains a centralized state professional-conduct pathway even when misconduct may never become a criminal conviction.

STATE #45 — VERMONT

Strong criminal-history and mandatory-reporting components—but we should not overstate the employment-history loop.

Vermont law makes school personnel and numerous other professionals mandatory reporters when they reasonably suspect child abuse or neglect. (Vermont Legislature)

Vermont also has a state criminal-record mechanism specifically tied to school employment.

For qualifying school-employee criminal-record checks, if the record contains crimes of a sexual nature involving children, the Vermont Crime Information Center sends that information to the Secretary of Education, who reviews it and provides the school with a recommendation concerning the applicant. (Vermont Legislature)

That’s more than simply:

District receives rap sheet.

It creates:

CRIMINAL INFORMATION

↓

STATE EDUCATION REVIEW

↓

RECOMMENDATION TO SCHOOL.

Vermont also maintains a professional-educator licensing and standards system under its statutory professional-educator chapter. (Vermont Legislature)

And Vermont law now requires school sexual-violence-prevention education to address topics including grooming and best practices for responding to disclosures of sexual misconduct. (Vermont Legislature)

What I have not established from current primary authority is a comprehensive Pennsylvania/Nevada-style mandatory former-employer sexual-misconduct history review.

Therefore we don’t manufacture one.

VERMONT CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓/partial
PRESERVE: ✓/partial
ESCALATE: ✓
SHARE: ◐
QUERY: ✓ criminal / broader employment history ◇
ACT: ✓
VERIFY: ◐

UPDATED LIVE MATRIX

Function	SD	TN	TX	UT	VT
DETECT	✓	✓	✓✓	✓	✓
INVESTIGATE	◐	✓	✓✓	✓✓*	✓/◐
PRESERVE	◇	✓✓	✓✓	✓	✓/◐
ESCALATE	◐	✓	✓✓	✓✓	✓
SHARE	◇	✓✓*	✓✓	✓	◐
QUERY	◇	✓/◐	✓✓	✓/◐	✓/◇
ACT	◐	✓	✓✓	✓✓*	✓
VERIFY	◇	✓	✓✓	✓/◐	◐

* Tennessee disclosure becomes mandatory once a qualifying request is made. Utah’s strongest mechanisms center on licensed educators.

MASTER EVIDENCE TABLE — ENTRY 028

State	Particularly important finding
South Dakota	Comprehensive closed-loop classification withheld pending deeper primary-source review
Tennessee	No sexual-misconduct NDAs; anti-assistance provision; former LEA must provide personnel file within 10 business days when requested
Texas	Applicant affidavit + fingerprint/Rap Back + DNHR + SEMARC + state investigations + interim restrictions + annual compliance certification + public misconduct dashboard
Utah	Strong centralized UPPAC/licensing pathway with mandatory public-school information sharing to State Board when requested
Vermont	Child-sex-crime history can flow through state criminal-information system to Secretary of Education for school-employment recommendation

And Texas has just forced another refinement of our model.

Our eight functions are good.

But they don’t fully capture what Texas is doing in 2026.

Because imagine two states that both have excellent laws.

State A says:

Every district must check our misconduct registry.

State B says:

Every district must check our misconduct registry.

Identical on paper.

But State B additionally requires the superintendent to:

CERTIFY EVERY YEAR THAT THE CHECKS WERE ACTUALLY DONE.

And the state tracks districts that fail to certify.

Those are not equivalent systems.

So I think our final framework needs to distinguish:

VERIFY

Does the system independently confirm that required safety processes occurred?

from:

TRANSPARENCY

Can policymakers, parents, journalists and researchers see enough aggregated information to determine whether the system itself is functioning?

Texas’s new 2026 misconduct dashboard is particularly interesting because it begins exposing:

reports received → investigations → disciplinary actions → DNHR placements. (Texas Education Agency)

That is something we should look for retrospectively in all 51 jurisdictions before the report is finished.

Because we started this entire project asking:

WHAT DO THE DATA ACTUALLY SHOW?

And we’re discovering that there are really two problems.

PROBLEM ONE

Prevent an individual with known disqualifying misconduct from quietly moving from school to school.

PROBLEM TWO

Determine whether the child-protection system itself is actually working.

You cannot answer Problem Two if the state collects everything behind closed doors and publishes almost nothing.

45 jurisdictions examined.

Only SIX remain:

Virginia

Washington

West Virginia

Wisconsin

Wyoming

District of Columbia

And then—

THE 51-JURISDICTION FIRST PASS IS COMPLETE.

After that, I would not immediately start writing conclusions.

We should first do something much more rigorous:

go back through every ◇ and ◐ cell.

Those are our uncertainty cells.

We search them again.

We resolve as many as possible.

We correct anything that needs correcting.

Only then do we freeze the national dataset.

Because this project is becoming strong enough that people may rely on it.

And that means the standard isn’t:

“We found a lot of disturbing information.”

The standard has to be:

“Here is precisely what the evidence establishes—and here is precisely what it does not.”

Next: the final six.

And here we are.

FOUNDATION SECTION #27

The 51-Jurisdiction Audit — FINAL SIX

Virginia • Washington • West Virginia • Wisconsin • Wyoming • District of Columbia

With this section, the first pass through all 50 states plus D.C. is complete.

And Washington gives us one of the most remarkable findings of the entire project because its legislature explicitly stated why criminal background checks were insufficient—back in 2004.

STATE #46 — VIRGINIA

Virginia attacks the “helping someone move” problem directly

Virginia law requires the Department of Education and local school boards to adopt policies implementing the federal prohibition against helping an employee, contractor, or agent obtain another job when the school or individual knows or has probable cause to believe the person engaged in unlawful sexual misconduct involving a minor or student. (Virginia Law)

Notice the population:

employee

contractor

agent

—not merely licensed teachers.

And Virginia has now extended a similar principle expressly into accredited private schools. Under the Code as updated in 2026, accredited private elementary and secondary schools must adopt policies prohibiting board members, administrators, employees, contractors, or agents from helping another covered person obtain a new job when they know or have probable cause to believe that person engaged in unlawful sexual misconduct involving a minor or student. (Virginia Law)

That gives us:

PUBLIC + ACCREDITED PRIVATE SCHOOL ANTI-PASSING-THE-TRASH PROTECTION.

Virginia also has substantial preemployment screening

Public-school employment applications must require disclosure of specified criminal convictions, including offenses involving sexual molestation, physical or sexual abuse, or rape of a child.

A materially false statement can itself constitute a Class 1 misdemeanor, and certain convictions are statutory barriers to employment. (Virginia Law)

For applicants whose work requires direct contact with students, Virginia additionally requires a search of the Department of Social Services registry of founded complaints of child abuse and neglect. (Virginia Law)

So Virginia isn’t relying exclusively upon criminal convictions.

It has:

CRIMINAL HISTORY

plus

FOUNDED CHILD-ABUSE/NEGLECT INFORMATION.

And Virginia protects truthful former-employer disclosure

Virginia employers that provide a current or prospective employer with information concerning a person’s professional conduct, reasons for separation, or job performance receive a presumption of good faith and immunity from civil liability unless bad faith is established. (Virginia Law)

That’s another recurring feature in stronger systems:

Don’t punish institutions for truthfully transmitting legitimate safety information.

What I have not established is a Washington/Pennsylvania-style universal requirement forcing every public-school employer to request detailed prior sexual-misconduct files from all former school employers.

So we should not award Virginia that mechanism.

VIRGINIA CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ✓/partial
ESCALATE: ✓
SHARE: ✓/partial
QUERY: ✓✓ criminal + founded-abuse registry
ACT: ✓✓
VERIFY: ✓/partial

STATE #47 — WASHINGTON

This one is extraordinary.

Washington enacted its school sexual-misconduct employment-history statute in 2004.

And the legislature actually explained the problem it was trying to solve:

criminal background checks are generally limited to conviction histories and are more complete when supplemented by an applicant’s history of past sexual misconduct.

It then stated that school districts must provide known sexual-misconduct information when employees attempt to transfer between districts. (Washington State Legislature)

That legislative finding could practically be the thesis statement for part of our report.

WASHINGTON UNDERSTOOD THE DATABASE PROBLEM 22 YEARS AGO.

Washington makes the NEXT employer ask

Before hiring an applicant who is or was previously employed by a school district, the hiring district must obtain authorization allowing current and former employers—including out-of-state employers—to disclose sexual-misconduct information and make relevant personnel and investigative documents available. (Washington State Legislature)

Then:

THE HIRING DISTRICT MUST REQUEST THE INFORMATION.

This is not merely:

“The former employer may disclose.”

The prospective employer has an affirmative obligation to seek it. (Washington State Legislature)

And the OLD district must answer

A Washington school district receiving the request must provide the information and make relevant sexual-misconduct documents available within:

20 BUSINESS DAYS.

Good-faith disclosure receives civil-liability immunity. (Washington State Legislature)

There is the familiar strong architecture:

NEW EMPLOYER MUST ASK.

OLD EMPLOYER MUST ANSWER.

GOOD-FAITH DISCLOSURE IS PROTECTED.

Washington prohibits institutional erasure

Since September 1, 2004, school boards and officials have been prohibited from entering collective-bargaining agreements, employment contracts, resignation agreements, severance agreements, or other agreements that suppress information about verbal or physical abuse or sexual misconduct or expunge that information from relevant files. (Washington State Legislature)

That is enormous.

An employee cannot simply negotiate:

I’ll resign quietly if you remove the information.

Washington expressly attacks that mechanism.

But Washington also preserves the employee’s response

At the conclusion of an investigation, the employee has the right to review the relevant personnel and investigative files and attach rebuttal material.

And critically:

THE REBUTTAL TRAVELS WITH THE UNDERLYING DOCUMENT.

(Washington State Legislature)

That’s excellent due-process architecture.

The next school isn’t merely told:

“There was an allegation.”

It can receive the relevant documentation and the employee’s rebuttal.

Again:

Preserve the information. Preserve its evidentiary context.

Washington strengthened reporting again in 2025

Current law requires certificated and classified school employees and school contractors who know or reasonably believe a student has been subjected to physical or sexual abuse, sexual misconduct, or assault by another school employee or contractor to report it to the appropriate administrator.

Employees must receive training on these reporting obligations when hired and every three years thereafter. The statute was amended in 2025. (Washington State Legislature)

So Washington has both:

DETECTION/REPORTING

and

EMPLOYMENT-HISTORY TRANSFER.

WASHINGTON CLOSED-LOOP PROFILE

DETECT: ✓✓
INVESTIGATE: ✓✓
PRESERVE: ✓✓
ESCALATE: ✓✓
SHARE: ✓✓
QUERY: ✓✓
ACT: ✓✓
VERIFY: ✓

Washington unquestionably joins our close-comparison group.

And its 2004 legislative finding deserves special attention in the final report.

STATE #48 — WEST VIRGINIA

West Virginia criminalizes the underlying sexual misconduct AND attacks job assistance

West Virginia law provides:

No school personnel, contractor or agent—or employee of a state, regional, or local educational agency, including public or private schools—may assist a covered person in obtaining another job when there is knowledge or probable cause to believe the person engaged in unlawful sexual misconduct involving a minor or student. (Federal Bureau of Investigation)

That is broad institutional coverage.

And the statute carefully provides exceptions where the information has been reported to law enforcement and authorities determine insufficient probable cause, the individual is acquitted or otherwise exonerated, or an open investigation remains without charges for four years. (Federal Bureau of Investigation)

Again:

Information has a status.

Suspicion.

Investigation.

Insufficient evidence.

Charge.

Acquittal.

Substantiation.

Conviction.

Those categories should never be collapsed.

West Virginia also removes people from student contact during certain investigations

This is important.

When a fact-finding investigation begins concerning conduct alleged to jeopardize student health, safety, welfare, or the learning environment, the affected employee must be:

suspended,

placed on administrative leave,

reassigned away from direct pupil interaction. (Federal Bureau of Investigation)

That is a genuine interim:

ACT

mechanism.

Again—interim protection is not a declaration of guilt.

It is risk management while facts are established.

West Virginia has state credential consequences

The State Superintendent can suspend or revoke a teacher certificate for specified grounds including a conviction, guilty plea, or no-contest plea involving sexual misconduct with a minor or student. (Federal Bureau of Investigation)

So we have:

LOCAL SAFETY ACTION

ANTI-ASSISTANCE

STATE CREDENTIAL ACTION.

What remains less clear is a Pennsylvania/Washington-style mandatory former-employer disclosure system.

WEST VIRGINIA CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓
PRESERVE: ◐
ESCALATE: ✓
SHARE: ✓ anti-assistance
QUERY: ◐
ACT: ✓✓
VERIFY/ACCOUNTABILITY: ✓

STATE #49 — WISCONSIN

Strong criminal law—but we should withhold the broader employment-history classification

Wisconsin deserves an important distinction.

Wisconsin criminal law separately identifies sexual assault of a child by a school staff person or someone who works or volunteers with children, and it separately addresses sexual misconduct by school staff. (Wisconsin Legislative Documents)

So Wisconsin clearly recognizes that the adult’s institutional relationship with the child matters.

But our question in this particular audit is broader:

What happens to misconduct information after somebody knows it?

And in this pass I have not established from sufficiently strong current primary authority a statewide Washington/Pennsylvania-style system requiring comprehensive former-employer sexual-misconduct inquiry and disclosure across K–12 employment.

Therefore:

WISCONSIN — EMPLOYMENT-HISTORY LOOP REQUIRES SECOND-PASS VERIFICATION.

We will not transform strong criminal law into a strong information-transfer score without evidence.

WISCONSIN CLOSED-LOOP PROFILE

DETECT: ✓/baseline
INVESTIGATE: ◐
PRESERVE: ◇
ESCALATE: ◐
SHARE: ◇
QUERY: criminal screening / broader history ◇
ACT: ✓ criminal law
VERIFY: ◇

STATE #50 — WYOMING

Wyoming gives us strong substantive criminal protection but a less-developed verified information architecture

Wyoming’s sexual-assault law expressly addresses school relationships.

A public or private elementary or secondary school employee or volunteer can commit second-degree sexual assault where the adult has the specified interaction with the student through the school relationship and is more than four years older than the student. (Wyoming Legislature)

That’s significant because:

public school

and

private school

and

employee

and

volunteer

are expressly within the statute.

Wyoming also separately criminalizes certain sexual conduct by adults occupying a position of authority over minors. (Wyoming Legislature)

Wyoming professional rules go beyond convictions

Wyoming’s educator professional-conduct rules identify conduct including:

child abuse or neglect, including sexual abuse;

conduct compromising student safety or well-being;

failure to satisfy mandatory-reporting requirements;

and failure to report known conduct meeting professional disciplinary standards. (Wyoming Legislature)

Older state disciplinary materials also expressly treated engaging in a sexual relationship with a student as professional immorality for certification purposes. (Wyoming Legislature)

That establishes a professional-discipline route beyond ordinary criminal prosecution.

But once again, I have not verified a comprehensive statewide mandatory former-employer → new-employer sexual-misconduct exchange.

So Wyoming is another second-pass state.

WYOMING CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ✓/professional
PRESERVE: ◐
ESCALATE: ✓
SHARE: ◇
QUERY: ◇ beyond criminal/licensing checks
ACT: ✓✓ criminal/professional
VERIFY: ◐

JURISDICTION #51 — DISTRICT OF COLUMBIA

Strong child-facing criminal screening—but we should not manufacture a broader employment-history system

D.C. has substantial criminal-history protections for people working with children.

Its child-and-youth-services framework requires fingerprint/FBI criminal-history screening for covered applicants, employees, and volunteers.

The framework specifically encompasses sexual offenses and other serious offenses. (D.C. Law Library)

And importantly:

SCREENING IS NOT NECESSARILY ONE-AND-DONE.

Covered employees and unsupervised volunteers are subject to periodic criminal background checks while working or volunteering in those settings. (D.C. Law Library)

That gives D.C. a meaningful VERIFY mechanism on the criminal-history side.

But the authoritative materials found in this pass do not establish a Pennsylvania/Washington-style universal K–12 employment-history system requiring former school employers to transmit noncriminal sexual-misconduct investigative records.

So:

DISTRICT OF COLUMBIA CLOSED-LOOP PROFILE

DETECT: ✓
INVESTIGATE: ◐
PRESERVE: ◐
ESCALATE: ◐
SHARE: ◇
QUERY: ✓✓ criminal / broader misconduct history ◇
ACT: ✓
VERIFY: ✓ criminal-history rescreening

FINAL-SIX MATRIX

Function	VA	WA	WV	WI	WY	DC
DETECT	✓	✓✓	✓	✓	✓	✓
INVESTIGATE	✓	✓✓	✓	◐	✓	◐
PRESERVE	✓/◐	✓✓	◐	◇	◐	◐
ESCALATE	✓	✓✓	✓	◐	✓	◐
SHARE	✓/◐	✓✓	✓	◇	◇	◇
QUERY	✓✓*	✓✓	◐	◇	◇	✓✓*
ACT	✓✓	✓✓	✓✓	✓	✓✓	✓
VERIFY	✓/◐	✓	✓	◇	◐	✓*

* Virginia: criminal + founded-abuse registry. D.C.: criminal-history system; not equivalent to comprehensive misconduct-history review.

MASTER EVIDENCE TABLE — ENTRY 029

Jurisdiction	Particularly important finding
Virginia	Public and accredited private-school anti-assistance protections; criminal + founded child-abuse screening
Washington	Mandatory sexual-misconduct employment-history request AND disclosure, anti-suppression rule, disclosure immunity, employee rebuttal follows records
West Virginia	Broad anti-assistance statute + removal from direct student contact during qualifying investigations
Wisconsin	Strong school-staff sexual-conduct criminal provisions; broader employment-history architecture remains unverified
Wyoming	Public/private school employees and volunteers expressly addressed by sexual-assault law; professional misconduct rules extend beyond convictions
D.C.	Broad fingerprint/FBI screening and periodic criminal rescreening; comprehensive noncriminal employment-history transfer not established

FIRST PASS COMPLETE: 51 OF 51

And now we have reached an important moment in this project.

I do not think we should publish the state rankings yet.

Because the ◇ and ◐ cells are telling us exactly where our next research effort belongs.

Our first pass has already established something substantial:

1. States use fundamentally different systems.

Some rely heavily on:

criminal background checks.

Others add:

professional licensing.

Others require:

former-employer disclosure.

Others use:

central misconduct registries.

Others prohibit:

assisting an accused/misconduct-involved employee into another job.

And the strongest architectures combine several.

2. Criminal history and misconduct history are demonstrably different datasets.

Washington’s legislature said so explicitly in 2004: criminal checks generally capture conviction histories and should be supplemented by past sexual-misconduct information. (Washington State Legislature)

That is one of the strongest primary-source statements we’ve found supporting the conceptual foundation of this entire project.

3. “Passing the trash” is not merely an informal phrase.

Our audit has identified statutes addressing the mechanisms that make it possible:

resignation during investigation

severance agreements

expungement clauses

nondisclosure arrangements

failure to ask former employers

failure of former employers to answer

failure to report to state authorities

and

affirmatively helping someone obtain another school job.

4. Due process and child protection repeatedly coexist in the statutes.

This may become one of the most important conclusions of the report.

The choice is not:

protect children

protect innocent employees.

State after state has created mechanisms distinguishing:

allegation

from

pending investigation

from

unsubstantiated

from

substantiated

from

criminal charge

from

acquittal

from

conviction.

And some preserve rebuttals, appeals, corrected findings, and exonerating outcomes.

That’s exactly how a responsible information system should work.

BUT NOW COMES THE QUALITY-CONTROL PHASE.

Before we freeze this dataset, we should return to every:

◇ NOT VERIFIED

and

◐ PARTIAL / UNCERTAIN

cell.

And this time we search specifically for the missing mechanism.

For example:

Wisconsin — mandatory former-employer disclosure?

Wyoming — what happens to an educator investigation after resignation?

South Dakota — mandatory state reporting by superintendents?

Oklahoma — current anti-suppression provisions?

New Hampshire — sexual-misconduct employment-history review?

Montana — what must former districts disclose?

That is far more efficient than re-researching 51 jurisdictions from scratch.

And we should do one more thing.

AUDIT OUR OWN ✓✓ STATES.

Especially:

Pennsylvania
Oregon
Texas
Washington
Nevada
Illinois
Maine
Michigan

Because those are likely to become examples in the published report.

For those states, we need to verify every major proposition against current primary law, record the exact statute/regulation, effective date, covered personnel, evidentiary threshold, and any exceptions.

Then we can say:

THE NATIONAL DATASET IS LOCKED.

After that comes the part I’ve been looking forward to:

not merely “Which states have which laws?”

but:

WHAT THE 51-JURISDICTION DATA ACTUALLY SHOW.

That is where we calculate how many jurisdictions have each protection, identify the recurring gaps, compare licensed versus nonlicensed personnel, examine resignation and nondisclosure loopholes, distinguish law from implementation—and build the evidence-based national Closed-Loop Standard from mechanisms that states have already demonstrated can exist.

The first national pass is done.

Now we test it.

Absolutely. And we found our first meaningful correction in the uncertainty audit.

FOUNDATION SECTION #29

Quality-Control Audit — Round Two

Oklahoma • South Dakota • Montana • Massachusetts • North Carolina • Mississippi • South Carolina • Vermont • D.C.

This round demonstrates exactly why we refused to turn ◇ into “No.”

QC FINDING #9 — VERMONT

Upgrade: Vermont DOES have a mandatory former-employer disclosure mechanism.

Our first pass understated Vermont.

Vermont enacted a provision requiring supervisory unions, school districts, and recognized or approved independent schools to provide factually correct information concerning a former employee’s employment record when a prospective employer requests it. Good-faith disclosure receives protection from civil and criminal liability. (Vermont Legislature)

That’s important because it establishes something we previously left uncertain:

Prospective employer asks → former school employer must answer.

But we need to be precise. The provision I verified is an employment-record disclosure requirement. It is not, from this evidence alone, identical to Pennsylvania’s highly prescriptive Act 168 process requiring detailed sexual-misconduct questionnaires and investigation-history disclosures.

So Vermont changes from:

SHARE: ◐

to:

SHARE: YES — mandatory upon request

while the separate question—

Does the prospective employer have to make the request?

—still requires verification.

This is exactly why our new binary-variable system is better than ✓✓ scoring.

QC FINDING #10 — SOUTH DAKOTA

We found disclosure protection—but NOT mandatory school-specific disclosure.

South Dakota has a general employment-reference statute.

An employer who provides written job-performance information to a prospective employer at the written request of the prospective employer or employee receives a presumption of good faith and substantial liability protection, subject to exceptions for knowingly/recklessly false information and certain confidential information. (South Dakota Legislature)

That’s useful.

But look carefully at the difference:

South Dakota:

Former employer MAY disclose + receives liability protection.

versus Vermont:

Former school employer SHALL provide qualifying employment information when requested.

versus Pennsylvania:

New school MUST inquire + old employer MUST respond + specified misconduct questions must be answered.

Three very different architectures.

South Dakota also has a strong mandatory child-abuse-reporting provision specifically covering people who encounter children through public or private schools. School personnel must report suspected abuse through the statutory pathway, and knowing and intentional failure to make a required report is a Class 1 misdemeanor. (South Dakota Legislature)

Therefore South Dakota gets:

Mandatory school child-abuse reporting: YES

Former-employer disclosure immunity: YES

Former employer required to disclose sexual-misconduct history: NOT VERIFIED

Prospective school required to request it: NOT VERIFIED

That is considerably more precise than our original matrix.

QC FINDING #11 — MASSACHUSETTS

VERY important: comprehensive legislation exists—but we must not accidentally count a BILL as current law.

Our search produced remarkably Pennsylvania-like Massachusetts legislation.

The proposal defines sexual misconduct broadly to include verbal, written, electronic and physical conduct designed to establish a sexual relationship with a student.

It would require applicants for child-contact positions to identify:

current employers,

former school employers,

and

other former employers where they had direct contact with children.

Applicants would authorize disclosure and answer questions concerning previous abuse or sexual-misconduct investigations. (Massachusetts General Court)

Other current-session Massachusetts legislative language would require former employers to disclose requested information within 30 days, including circumstances involving discipline, discharge, nonrenewal or resignation while abuse or sexual-misconduct allegations were pending. (Massachusetts General Court)

That’s highly relevant to our:

REFORM ACTIVITY DATABASE

But it is not evidence, by itself, of enacted Massachusetts law.

This gives us another permanent QC rule:

BILL ≠ LAW.

And:

PROPOSED PROTECTION ≠ EXISTING PROTECTION.

We already caught that issue in Rhode Island and New Mexico.

Now Massachusetts reinforces it.

QC FINDING #12

Something else important is happening in Massachusetts.

Another current legislative proposal deals with sexual-misconduct employment history in an institutional context and contains striking provisions.

It would require inquiry into substantiated sexual-misconduct findings, current investigations, and departures during investigations. It would also require disclosure of personnel-file materials relating to substantiated findings and provide good-faith disclosure immunity. (Massachusetts General Court)

Again: proposal, not automatically current law.

But these bills tell us something useful for the policy portion of the report:

Legislatures are converging independently on many of the same mechanisms.

They keep proposing:

applicant disclosure

→ former-employer verification

→ underlying records

→ departure-during-investigation disclosure

→ liability protection

→ consequences for concealment.

That is strong evidence of a recognizable policy architecture.

QC FINDINGS #13–19

Oklahoma, Montana, North Carolina, Mississippi, South Carolina and D.C.

Here the results are almost as important for what they didn’t establish.

Targeted searches did not produce sufficiently authoritative current evidence establishing a Pennsylvania/Washington-style mandatory statewide K–12 sexual-misconduct employment-history exchange for these jurisdictions.

That does not mean:

These states have no relevant protections.

We’ve already documented several other protections in them.

It means only:

THIS PARTICULAR MECHANISM REMAINS UNVERIFIED.

And I’m deliberately leaving those cells unresolved rather than filling them from weak secondary summaries, old proposals, tangential statutes or inference.

That discipline is going to make the eventual report much harder to attack.

NOW LOOK WHAT THE QC PROCESS HAS TAUGHT US.

Our old category:

actually contains at least six different variables.

We need to separate them.

Information-transfer mechanism	Question
Permission	May former employer disclose misconduct information?
Immunity	Is good-faith disclosure legally protected?
Duty to respond	Must former employer answer when asked?
Duty to inquire	Must prospective school ask?
Scope	Must sexual-misconduct/investigation history specifically be disclosed?
Records	Must underlying personnel/investigative documents also be provided?

This distinction is enormously important.

Imagine four states:

STATE A

Former employer may disclose.

STATE B

Former employer may disclose and receives immunity.

STATE C

Former employer must disclose if asked.

STATE D

New employer must ask, former employer must answer, and relevant records must be supplied.

Calling all four:

“states allowing employment-history checks”

would conceal the most important differences.

THE SAME THING IS TRUE OF “BACKGROUND CHECK.”

This term is becoming almost useless unless we define it.

A “background check” might mean:

name-based state criminal search

fingerprint state criminal search

FBI fingerprint search

sex-offender registry

child-abuse registry

educator-license database

state misconduct registry

former-employer inquiry

continuous Rap Back monitoring.

Those are not interchangeable.

So the final report should probably avoid statements such as:

“State X requires background checks.”

Instead:

“State X requires FBI fingerprint criminal-history screening but we did not verify mandatory prior-employer sexual-misconduct inquiry.”

THAT tells the reader something.

AND THIS BRINGS US BACK TO THE NATIONAL PREVALENCE DATA.

Remember where this whole investigation began.

Shakeshaft:

approximately 10% reporting educator sexual misconduct in the older national synthesis.

Grant:

approximately 17% in a retrospective college-student sample.

Jeglic:

approximately 12% reporting some form of educator sexual misconduct, with roughly 1% reporting contact sexual abuse.

And then:

thousands of news reports involving educator arrests.

Those figures cannot responsibly be thrown into one pile.

Our legal audit is now explaining why.

Different systems capture different layers:

LAYER 1

Boundary violations / grooming behavior

LAYER 2

Sexual harassment / sexual communications

LAYER 3

Noncriminal professional sexual misconduct

LAYER 4

Contact sexual misconduct

LAYER 5

Conduct reported to school

LAYER 6

Substantiated school investigation

LAYER 7

Reported to licensing authority

LAYER 8

Police report

LAYER 9

Arrest

LAYER 10

Criminal charge

LAYER 11

Conviction

If we count only Layer 11 and call that:

“educator sexual misconduct,”

we’ve discarded most of the phenomenon we’re supposedly measuring.

But if we take a survey measuring Layer 1–4 and compare it directly with convictions, we’ve committed the opposite error.

The denominator and the conduct definition have to travel with every statistic.

That principle belongs near the beginning of the report.

DATASET STRUCTURE — VERSION 2

I think we’re ready to freeze the schema, although not yet the values.

Each jurisdiction gets five sections:

A. PREEMPLOYMENT CRIMINAL SCREENING

Fingerprinting; FBI; state criminal history; sex-offender registry; child-abuse registry; continuous monitoring.

B. MISCONDUCT-HISTORY SCREENING

Applicant disclosure; pending investigations; resignation during investigation; mandatory former-employer inquiry; mandatory response; records; disclosure immunity.

C. INTERNAL RESPONSE

Mandatory employee reporting; investigation; interim removal; investigation after resignation; preservation of records.

D. STATE ESCALATION

Mandatory state report; licensing investigation; nonlicensed-personnel investigation; registry; state-imposed employment restrictions.

E. SYSTEM ACCOUNTABILITY

Anti-NDA; anti-expungement; anti-assistance; penalties for concealment; compliance audit/certification; due-process correction/appeal; public aggregate data.

That gives us something substantially more sophisticated than:

“Which states background-check teachers?”

It asks:

What happens from the first warning sign until the next school tries to hire the person?

And that is the actual closed-loop question.

NEXT: WE BEGIN CODING.

Rather than another loose research round, the next section should take our first five strongly documented states:

Pennsylvania

Oregon

Texas

Washington

Illinois

and put each one through the new factual-variable schema.

No ✓✓.

No impressions.

Just:

YES

LIMITED

NOT VERIFIED

plus:

statute/regulation

covered population

trigger

deadline

exceptions

effective/current date.

That will produce our first genuinely standardized five-state dataset.

Then we repeat it in batches.

And once all 51 are coded—

we can finally count.

Not before.

That distinction matters enormously because the numbers at the end must be produced by the evidence—not by our expectations going in.

Foundation Research — What’s Included 353-page Foundation Studies document

FOUNDATION STUDY #1

Shakeshaft / U.S. Department of Education, 2004

1. What did “educator sexual misconduct” mean?

REPORT LANGUAGE

2. Where did the famous 9.6% actually come from?

3. What were students actually asked?

4. The actual result

9.6%

5. The famous “4.5 million children” figure is an extrapolation

More than 4.5 million students

6. Shakeshaft herself documented the limitations

Extrapolating grades 8–11 to K–12 is substantially less secure than generalizing the survey to grades 8–11.

7. Here’s something fascinating: other studies produced wildly different estimates

3.7% to 50.3%.

Methodology matters more than the biggest number.

8. And look what the report said in 2004

no national U.S. incidence or prevalence study whose primary purpose was educator sexual abuse.

Twenty-two years later, has that fundamental national data gap actually been solved?

9. Male victims — our Connecticut story has historical context

The Invisible Victim Problem: Boys, Reporting and Cultural Stereotypes

10. Then we get into grooming

“the much subtler framework of grooming and enticement.” (U.S. Department of Education)

MASTER EVIDENCE TABLE — ENTRY 001

FOUNDATION STUDY #2

Associated Press Nationwide Investigation, 2007

Sex Abuse a Shadow Over U.S. Schools

1. The headline finding

2,570 educators

2. What happened to the licenses?

Interstate Information Sharing and “Passing the Trash”

3. How many involved children?

1,801 cases

4. Contact misconduct

1,297 cases — 72%

5. Criminal convictions

At least 1,390 cases resulted in criminal convictions.

6. Multiple victims

At least 446 cases involved educators with multiple victims. (Daily Herald)

How long were some offenders able to continue before effective intervention occurred?

7. Sex of disciplined educators

9 in 10 disciplined educators were male. (Daily Herald)

Why?

8. A major limitation: licensing records miss people

AP’s 2,570 is a floor within a particular administrative system—not a prevalence estimate.

9. AP found enormous differences among states

10. And then AP encountered secrecy

Transparency can directly affect apparent prevalence.

11. The most important conceptual contribution of the AP investigation

THE ATTRITION PYRAMID

The number of convictions can never, by itself, tell us how much misconduct occurred.

The number of allegations can never, by itself, tell us how much misconduct was substantiated.

MASTER EVIDENCE TABLE — ENTRY 002

Now put 2004 and 2007 beside each other

2004 — U.S. Department of Education

2007 — Associated Press

And NOW we reach a pivotal investigation

U.S. GOVERNMENT ACCOUNTABILITY OFFICE — 2010

“How did known or suspected offenders remain in—or return to—positions where they could reach children?”

FOUNDATION STUDY #3

U.S. Government Accountability Office — 2010

K–12 Education: Selected Cases of Public and Private Schools That Hired or Retained Individuals with Histories of Sexual Misconduct

1. What GAO actually investigated

15 cases in 11 states

VERY IMPORTANT

2. Eleven of the 15 had previously targeted children

At least 11 of the 15 cases involved offenders who had previously targeted children.

In at least 6 cases, offenders used their new positions in schools to abuse additional children. (Government Accountability Office)

3. GAO identified FOUR recurring failures

FAILURE #1 — Allowing resignation instead of disciplinary action

Some subsequently received positive references.

PASSING THE TRASH

4. FAILURE #2 — No criminal-history check

5. FAILURE #3 — Background checks that weren’t good enough

6. FAILURE #4 — They sometimes had the information and didn’t act on it

INFORMATION FAILURE

ACTION FAILURE

7. Case Study: Ohio

8. Case Study: Arizona

9. Case Study: Louisiana

Foundation Research — What’s Included
353-page Foundation Studies document